maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/llvm/test/Bitcode/compatibility.ll

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

1930 lines
70 KiB
LLVM
Raw Normal View History

IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
; Bitcode compatibility test for llvm
;
; Please update this file when making any IR changes. Information on the
; release process for this file is available here:
;
Add speculatable function attribute This attribute tells the optimizer that the function may be speculated. Patch by Tom Stellard llvm-svn: 301680
2017-04-29 04:25:27 +08:00
; http://llvm.org/docs/DeveloperPolicy.html#ir-backwards-compatibility
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
; RUN: llvm-as < %s | llvm-dis | llvm-as | llvm-dis | FileCheck %s
[Bitcode][Asm] Teach LLVM to read and write operand bundles. Summary: This also adds the first set of tests for operand bundles. The optimizer has not been audited to ensure that it does the right thing with operand bundles. Depends on D12456. Reviewers: reames, chandlerc, majnemer, dexonsmith, kmod, JosephTremoulet, rnk, bogner Subscribers: maksfb, llvm-commits Differential Revision: http://reviews.llvm.org/D12457 llvm-svn: 248551
2015-09-25 07:34:52 +08:00
; RUN-PR24755: verify-uselistorder < %s
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
target datalayout = "E"
; CHECK: target datalayout = "E"
target triple = "x86_64-apple-macosx10.10.0"
; CHECK: target triple = "x86_64-apple-macosx10.10.0"
;; Module-level assembly
module asm "beep boop"
; CHECK: module asm "beep boop"
;; Comdats
$comdat.any = comdat any
; CHECK: $comdat.any = comdat any
$comdat.exactmatch = comdat exactmatch
; CHECK: $comdat.exactmatch = comdat exactmatch
$comdat.largest = comdat largest
; CHECK: $comdat.largest = comdat largest
$comdat.noduplicates = comdat noduplicates
; CHECK: $comdat.noduplicates = comdat noduplicates
$comdat.samesize = comdat samesize
; CHECK: $comdat.samesize = comdat samesize
;; Constants
@const.true = constant i1 true
; CHECK: @const.true = constant i1 true
@const.false = constant i1 false
; CHECK: @const.false = constant i1 false
@const.int = constant i32 zeroinitializer
; CHECK: @const.int = constant i32 0
@const.float = constant double 0.0
; CHECK: @const.float = constant double 0.0
@const.null = constant i8* null
; CHECK: @const.null = constant i8* null
%const.struct.type = type { i32, i8 }
%const.struct.type.packed = type <{ i32, i8 }>
@const.struct = constant %const.struct.type { i32 -1, i8 undef }
; CHECK: @const.struct = constant %const.struct.type { i32 -1, i8 undef }
@const.struct.packed = constant %const.struct.type.packed <{ i32 -1, i8 1 }>
; CHECK: @const.struct.packed = constant %const.struct.type.packed <{ i32 -1, i8 1 }>
Bitcode: Move these tests into compatibility.ll I added a couple of tests in r256982, but vedantk suggested that they fit better into compatibility.ll, since they could catch format breaks later on there. llvm-svn: 256990
2016-01-07 07:16:37 +08:00
; CHECK: @constant.array.i8 = constant [3 x i8] c"\00\01\00"
@constant.array.i8 = constant [3 x i8] [i8 -0, i8 1, i8 0]
; CHECK: @constant.array.i16 = constant [3 x i16] [i16 0, i16 1, i16 0]
@constant.array.i16 = constant [3 x i16] [i16 -0, i16 1, i16 0]
; CHECK: @constant.array.i32 = constant [3 x i32] [i32 0, i32 1, i32 0]
@constant.array.i32 = constant [3 x i32] [i32 -0, i32 1, i32 0]
; CHECK: @constant.array.i64 = constant [3 x i64] [i64 0, i64 1, i64 0]
@constant.array.i64 = constant [3 x i64] [i64 -0, i64 1, i64 0]
; CHECK: @constant.array.f16 = constant [3 x half] [half 0xH8000, half 0xH3C00, half 0xH0000]
@constant.array.f16 = constant [3 x half] [half -0.0, half 1.0, half 0.0]
; CHECK: @constant.array.f32 = constant [3 x float] [float -0.000000e+00, float 1.000000e+00, float 0.000000e+00]
@constant.array.f32 = constant [3 x float] [float -0.0, float 1.0, float 0.0]
; CHECK: @constant.array.f64 = constant [3 x double] [double -0.000000e+00, double 1.000000e+00, double 0.000000e+00]
@constant.array.f64 = constant [3 x double] [double -0.0, double 1.0, double 0.0]
; CHECK: @constant.vector.i8 = constant <3 x i8> <i8 0, i8 1, i8 0>
@constant.vector.i8 = constant <3 x i8> <i8 -0, i8 1, i8 0>
; CHECK: @constant.vector.i16 = constant <3 x i16> <i16 0, i16 1, i16 0>
@constant.vector.i16 = constant <3 x i16> <i16 -0, i16 1, i16 0>
; CHECK: @constant.vector.i32 = constant <3 x i32> <i32 0, i32 1, i32 0>
@constant.vector.i32 = constant <3 x i32> <i32 -0, i32 1, i32 0>
; CHECK: @constant.vector.i64 = constant <3 x i64> <i64 0, i64 1, i64 0>
@constant.vector.i64 = constant <3 x i64> <i64 -0, i64 1, i64 0>
; CHECK: @constant.vector.f16 = constant <3 x half> <half 0xH8000, half 0xH3C00, half 0xH0000>
@constant.vector.f16 = constant <3 x half> <half -0.0, half 1.0, half 0.0>
; CHECK: @constant.vector.f32 = constant <3 x float> <float -0.000000e+00, float 1.000000e+00, float 0.000000e+00>
@constant.vector.f32 = constant <3 x float> <float -0.0, float 1.0, float 0.0>
; CHECK: @constant.vector.f64 = constant <3 x double> <double -0.000000e+00, double 1.000000e+00, double 0.000000e+00>
@constant.vector.f64 = constant <3 x double> <double -0.0, double 1.0, double 0.0>
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
;; Global Variables
; Format: [@<GlobalVarName> =] [Linkage] [Visibility] [DLLStorageClass]
IR: Introduce local_unnamed_addr attribute. If a local_unnamed_addr attribute is attached to a global, the address is known to be insignificant within the module. It is distinct from the existing unnamed_addr attribute in that it only describes a local property of the module rather than a global property of the symbol. This attribute is intended to be used by the code generator and LTO to allow the linker to decide whether the global needs to be in the symbol table. It is possible to exclude a global from the symbol table if three things are true: - This attribute is present on every instance of the global (which means that the normal rule that the global must have a unique address can be broken without being observable by the program by performing comparisons against the global's address) - The global has linkonce_odr linkage (which means that each linkage unit must have its own copy of the global if it requires one, and the copy in each linkage unit must be the same) - It is a constant or a function (which means that the program cannot observe that the unique-address rule has been broken by writing to the global) Although this attribute could in principle be computed from the module contents, LTO clients (i.e. linkers) will normally need to be able to compute this property as part of symbol resolution, and it would be inefficient to materialize every module just to compute it. See: http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20160509/356401.html http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20160516/356738.html for earlier discussion. Part of the fix for PR27553. Differential Revision: http://reviews.llvm.org/D20348 llvm-svn: 272709
2016-06-15 05:01:22 +08:00
; [ThreadLocal] [(unnamed_addr|local_unnamed_addr)] [AddrSpace] [ExternallyInitialized]
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
; <global | constant> <Type> [<InitializerConstant>]
; [, section "name"] [, comdat [($name)]] [, align <Alignment>]
; Global Variables -- Simple
@g1 = global i32 0
; CHECK: @g1 = global i32 0
@g2 = constant i32 0
; CHECK: @g2 = constant i32 0
; Global Variables -- Linkage
@g.private = private global i32 0
; CHECK: @g.private = private global i32 0
@g.internal = internal global i32 0
; CHECK: @g.internal = internal global i32 0
@g.available_externally = available_externally global i32 0
; CHECK: @g.available_externally = available_externally global i32 0
@g.linkonce = linkonce global i32 0
; CHECK: @g.linkonce = linkonce global i32 0
@g.weak = weak global i32 0
; CHECK: @g.weak = weak global i32 0
@g.common = common global i32 0
; CHECK: @g.common = common global i32 0
@g.appending = appending global [4 x i8] c"test"
; CHECK: @g.appending = appending global [4 x i8] c"test"
@g.extern_weak = extern_weak global i32
; CHECK: @g.extern_weak = extern_weak global i32
@g.linkonce_odr = linkonce_odr global i32 0
; CHECK: @g.linkonce_odr = linkonce_odr global i32 0
@g.weak_odr = weak_odr global i32 0
; CHECK: @g.weak_odr = weak_odr global i32 0
@g.external = external global i32
; CHECK: @g.external = external global i32
; Global Variables -- Visibility
@g.default = default global i32 0
; CHECK: @g.default = global i32 0
@g.hidden = hidden global i32 0
; CHECK: @g.hidden = hidden global i32 0
@g.protected = protected global i32 0
; CHECK: @g.protected = protected global i32 0
; Global Variables -- DLLStorageClass
@g.dlldefault = default global i32 0
; CHECK: @g.dlldefault = global i32 0
@g.dllimport = external dllimport global i32
; CHECK: @g.dllimport = external dllimport global i32
@g.dllexport = dllexport global i32 0
; CHECK: @g.dllexport = dllexport global i32 0
; Global Variables -- ThreadLocal
@g.notthreadlocal = global i32 0
; CHECK: @g.notthreadlocal = global i32 0
@g.generaldynamic = thread_local global i32 0
; CHECK: @g.generaldynamic = thread_local global i32 0
@g.localdynamic = thread_local(localdynamic) global i32 0
; CHECK: @g.localdynamic = thread_local(localdynamic) global i32 0
@g.initialexec = thread_local(initialexec) global i32 0
; CHECK: @g.initialexec = thread_local(initialexec) global i32 0
@g.localexec = thread_local(localexec) global i32 0
; CHECK: @g.localexec = thread_local(localexec) global i32 0
IR: Introduce local_unnamed_addr attribute. If a local_unnamed_addr attribute is attached to a global, the address is known to be insignificant within the module. It is distinct from the existing unnamed_addr attribute in that it only describes a local property of the module rather than a global property of the symbol. This attribute is intended to be used by the code generator and LTO to allow the linker to decide whether the global needs to be in the symbol table. It is possible to exclude a global from the symbol table if three things are true: - This attribute is present on every instance of the global (which means that the normal rule that the global must have a unique address can be broken without being observable by the program by performing comparisons against the global's address) - The global has linkonce_odr linkage (which means that each linkage unit must have its own copy of the global if it requires one, and the copy in each linkage unit must be the same) - It is a constant or a function (which means that the program cannot observe that the unique-address rule has been broken by writing to the global) Although this attribute could in principle be computed from the module contents, LTO clients (i.e. linkers) will normally need to be able to compute this property as part of symbol resolution, and it would be inefficient to materialize every module just to compute it. See: http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20160509/356401.html http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20160516/356738.html for earlier discussion. Part of the fix for PR27553. Differential Revision: http://reviews.llvm.org/D20348 llvm-svn: 272709
2016-06-15 05:01:22 +08:00
; Global Variables -- unnamed_addr and local_unnamed_addr
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
@g.unnamed_addr = unnamed_addr global i32 0
; CHECK: @g.unnamed_addr = unnamed_addr global i32 0
IR: Introduce local_unnamed_addr attribute. If a local_unnamed_addr attribute is attached to a global, the address is known to be insignificant within the module. It is distinct from the existing unnamed_addr attribute in that it only describes a local property of the module rather than a global property of the symbol. This attribute is intended to be used by the code generator and LTO to allow the linker to decide whether the global needs to be in the symbol table. It is possible to exclude a global from the symbol table if three things are true: - This attribute is present on every instance of the global (which means that the normal rule that the global must have a unique address can be broken without being observable by the program by performing comparisons against the global's address) - The global has linkonce_odr linkage (which means that each linkage unit must have its own copy of the global if it requires one, and the copy in each linkage unit must be the same) - It is a constant or a function (which means that the program cannot observe that the unique-address rule has been broken by writing to the global) Although this attribute could in principle be computed from the module contents, LTO clients (i.e. linkers) will normally need to be able to compute this property as part of symbol resolution, and it would be inefficient to materialize every module just to compute it. See: http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20160509/356401.html http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20160516/356738.html for earlier discussion. Part of the fix for PR27553. Differential Revision: http://reviews.llvm.org/D20348 llvm-svn: 272709
2016-06-15 05:01:22 +08:00
@g.local_unnamed_addr = local_unnamed_addr global i32 0
; CHECK: @g.local_unnamed_addr = local_unnamed_addr global i32 0
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
; Global Variables -- AddrSpace
@g.addrspace = addrspace(1) global i32 0
; CHECK: @g.addrspace = addrspace(1) global i32 0
; Global Variables -- ExternallyInitialized
@g.externally_initialized = external externally_initialized global i32
; CHECK: @g.externally_initialized = external externally_initialized global i32
; Global Variables -- section
@g.section = global i32 0, section "_DATA"
; CHECK: @g.section = global i32 0, section "_DATA"
Add IR support, ELF section and user documentation for partitioning feature. The partitioning feature was proposed here: http://lists.llvm.org/pipermail/llvm-dev/2019-February/130583.html This is mostly just documentation. The feature itself will be contributed in subsequent patches. Differential Revision: https://reviews.llvm.org/D60242 llvm-svn: 361923
2019-05-29 11:29:01 +08:00
; Global Variables -- partition
@g.partition = global i32 0, partition "part"
; CHECK: @g.partition = global i32 0, partition "part"
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
; Global Variables -- comdat
@comdat.any = global i32 0, comdat
; CHECK: @comdat.any = global i32 0, comdat
@comdat.exactmatch = global i32 0, comdat
; CHECK: @comdat.exactmatch = global i32 0, comdat
@comdat.largest = global i32 0, comdat
; CHECK: @comdat.largest = global i32 0, comdat
@comdat.noduplicates = global i32 0, comdat
; CHECK: @comdat.noduplicates = global i32 0, comdat
@comdat.samesize = global i32 0, comdat
; CHECK: @comdat.samesize = global i32 0, comdat
; Force two globals from different comdats into sections with the same name.
$comdat1 = comdat any
$comdat2 = comdat any
@g.comdat1 = global i32 0, section "SharedSection", comdat($comdat1)
; CHECK: @g.comdat1 = global i32 0, section "SharedSection", comdat($comdat1)
@g.comdat2 = global i32 0, section "SharedSection", comdat($comdat2)
; CHECK: @g.comdat2 = global i32 0, section "SharedSection", comdat($comdat2)
; Global Variables -- align
@g.align = global i32 0, align 4
; CHECK: @g.align = global i32 0, align 4
; Global Variables -- Intrinsics
%pri.func.data = type { i32, void ()*, i8* }
@g.used1 = global i32 0
@g.used2 = global i32 0
@g.used3 = global i8 0
declare void @g.f1()
@llvm.used = appending global [1 x i32*] [i32* @g.used1], section "llvm.metadata"
; CHECK: @llvm.used = appending global [1 x i32*] [i32* @g.used1], section "llvm.metadata"
@llvm.compiler.used = appending global [1 x i32*] [i32* @g.used2], section "llvm.metadata"
; CHECK: @llvm.compiler.used = appending global [1 x i32*] [i32* @g.used2], section "llvm.metadata"
@llvm.global_ctors = appending global [1 x %pri.func.data] [%pri.func.data { i32 0, void ()* @g.f1, i8* @g.used3 }], section "llvm.metadata"
; CHECK: @llvm.global_ctors = appending global [1 x %pri.func.data] [%pri.func.data { i32 0, void ()* @g.f1, i8* @g.used3 }], section "llvm.metadata"
@llvm.global_dtors = appending global [1 x %pri.func.data] [%pri.func.data { i32 0, void ()* @g.f1, i8* @g.used3 }], section "llvm.metadata"
; CHECK: @llvm.global_dtors = appending global [1 x %pri.func.data] [%pri.func.data { i32 0, void ()* @g.f1, i8* @g.used3 }], section "llvm.metadata"
;; Aliases
; Format: @<Name> = [Linkage] [Visibility] [DLLStorageClass] [ThreadLocal]
; [unnamed_addr] alias <AliaseeTy> @<Aliasee>
; Aliases -- Linkage
[opaque pointer type] Add textual IR support for explicit type parameter for global aliases update.py: import fileinput import sys import re alias_match_prefix = r"(.*(?:=|:|^)\s*(?:external |)(?:(?:private|internal|linkonce|linkonce_odr|weak|weak_odr|common|appending|extern_weak|available_externally) )?(?:default |hidden |protected )?(?:dllimport |dllexport )?(?:unnamed_addr |)(?:thread_local(?:\([a-z]*\))? )?alias" plain = re.compile(alias_match_prefix + r" (.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|addrspacecast|\[\[[a-zA-Z]|\{\{).*$)") cast = re.compile(alias_match_prefix + r") ((?:bitcast|inttoptr|addrspacecast)\s*\(.* to (.*?)(| addrspace\(\d+\) *)\*\)\s*(?:;.*)?$)") gep = re.compile(alias_match_prefix + r") ((?:getelementptr)\s*(?:inbounds)?\s*\((?P<type>.*), (?P=type)(?:\s*addrspace\(\d+\)\s*)?\* .*\)\s*(?:;.*)?$)") def conv(line): m = re.match(cast, line) if m: return m.group(1) + " " + m.group(3) + ", " + m.group(2) m = re.match(gep, line) if m: return m.group(1) + " " + m.group(3) + ", " + m.group(2) m = re.match(plain, line) if m: return m.group(1) + ", " + m.group(2) + m.group(3) + "*" + m.group(4) + "\n" return line for line in sys.stdin: sys.stdout.write(conv(line)) apply.sh: for name in "$@" do python3 `dirname "$0"`/update.py < "$name" > "$name.tmp" && mv "$name.tmp" "$name" rm -f "$name.tmp" done The actual commands: From llvm/src: find test/ -name *.ll | xargs ./apply.sh From llvm/src/tools/clang: find test/ -name *.mm -o -name *.m -o -name *.cpp -o -name *.c | xargs -I '{}' ../../apply.sh "{}" From llvm/src/tools/polly: find test/ -name *.ll | xargs ./apply.sh llvm-svn: 247378
2015-09-11 11:22:04 +08:00
@a.private = private alias i32, i32* @g.private
; CHECK: @a.private = private alias i32, i32* @g.private
@a.internal = internal alias i32, i32* @g.internal
; CHECK: @a.internal = internal alias i32, i32* @g.internal
@a.linkonce = linkonce alias i32, i32* @g.linkonce
; CHECK: @a.linkonce = linkonce alias i32, i32* @g.linkonce
@a.weak = weak alias i32, i32* @g.weak
; CHECK: @a.weak = weak alias i32, i32* @g.weak
@a.linkonce_odr = linkonce_odr alias i32, i32* @g.linkonce_odr
; CHECK: @a.linkonce_odr = linkonce_odr alias i32, i32* @g.linkonce_odr
@a.weak_odr = weak_odr alias i32, i32* @g.weak_odr
; CHECK: @a.weak_odr = weak_odr alias i32, i32* @g.weak_odr
@a.external = external alias i32, i32* @g1
; CHECK: @a.external = alias i32, i32* @g1
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
; Aliases -- Visibility
[opaque pointer type] Add textual IR support for explicit type parameter for global aliases update.py: import fileinput import sys import re alias_match_prefix = r"(.*(?:=|:|^)\s*(?:external |)(?:(?:private|internal|linkonce|linkonce_odr|weak|weak_odr|common|appending|extern_weak|available_externally) )?(?:default |hidden |protected )?(?:dllimport |dllexport )?(?:unnamed_addr |)(?:thread_local(?:\([a-z]*\))? )?alias" plain = re.compile(alias_match_prefix + r" (.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|addrspacecast|\[\[[a-zA-Z]|\{\{).*$)") cast = re.compile(alias_match_prefix + r") ((?:bitcast|inttoptr|addrspacecast)\s*\(.* to (.*?)(| addrspace\(\d+\) *)\*\)\s*(?:;.*)?$)") gep = re.compile(alias_match_prefix + r") ((?:getelementptr)\s*(?:inbounds)?\s*\((?P<type>.*), (?P=type)(?:\s*addrspace\(\d+\)\s*)?\* .*\)\s*(?:;.*)?$)") def conv(line): m = re.match(cast, line) if m: return m.group(1) + " " + m.group(3) + ", " + m.group(2) m = re.match(gep, line) if m: return m.group(1) + " " + m.group(3) + ", " + m.group(2) m = re.match(plain, line) if m: return m.group(1) + ", " + m.group(2) + m.group(3) + "*" + m.group(4) + "\n" return line for line in sys.stdin: sys.stdout.write(conv(line)) apply.sh: for name in "$@" do python3 `dirname "$0"`/update.py < "$name" > "$name.tmp" && mv "$name.tmp" "$name" rm -f "$name.tmp" done The actual commands: From llvm/src: find test/ -name *.ll | xargs ./apply.sh From llvm/src/tools/clang: find test/ -name *.mm -o -name *.m -o -name *.cpp -o -name *.c | xargs -I '{}' ../../apply.sh "{}" From llvm/src/tools/polly: find test/ -name *.ll | xargs ./apply.sh llvm-svn: 247378
2015-09-11 11:22:04 +08:00
@a.default = default alias i32, i32* @g.default
; CHECK: @a.default = alias i32, i32* @g.default
@a.hidden = hidden alias i32, i32* @g.hidden
; CHECK: @a.hidden = hidden alias i32, i32* @g.hidden
@a.protected = protected alias i32, i32* @g.protected
; CHECK: @a.protected = protected alias i32, i32* @g.protected
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
; Aliases -- DLLStorageClass
[opaque pointer type] Add textual IR support for explicit type parameter for global aliases update.py: import fileinput import sys import re alias_match_prefix = r"(.*(?:=|:|^)\s*(?:external |)(?:(?:private|internal|linkonce|linkonce_odr|weak|weak_odr|common|appending|extern_weak|available_externally) )?(?:default |hidden |protected )?(?:dllimport |dllexport )?(?:unnamed_addr |)(?:thread_local(?:\([a-z]*\))? )?alias" plain = re.compile(alias_match_prefix + r" (.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|addrspacecast|\[\[[a-zA-Z]|\{\{).*$)") cast = re.compile(alias_match_prefix + r") ((?:bitcast|inttoptr|addrspacecast)\s*\(.* to (.*?)(| addrspace\(\d+\) *)\*\)\s*(?:;.*)?$)") gep = re.compile(alias_match_prefix + r") ((?:getelementptr)\s*(?:inbounds)?\s*\((?P<type>.*), (?P=type)(?:\s*addrspace\(\d+\)\s*)?\* .*\)\s*(?:;.*)?$)") def conv(line): m = re.match(cast, line) if m: return m.group(1) + " " + m.group(3) + ", " + m.group(2) m = re.match(gep, line) if m: return m.group(1) + " " + m.group(3) + ", " + m.group(2) m = re.match(plain, line) if m: return m.group(1) + ", " + m.group(2) + m.group(3) + "*" + m.group(4) + "\n" return line for line in sys.stdin: sys.stdout.write(conv(line)) apply.sh: for name in "$@" do python3 `dirname "$0"`/update.py < "$name" > "$name.tmp" && mv "$name.tmp" "$name" rm -f "$name.tmp" done The actual commands: From llvm/src: find test/ -name *.ll | xargs ./apply.sh From llvm/src/tools/clang: find test/ -name *.mm -o -name *.m -o -name *.cpp -o -name *.c | xargs -I '{}' ../../apply.sh "{}" From llvm/src/tools/polly: find test/ -name *.ll | xargs ./apply.sh llvm-svn: 247378
2015-09-11 11:22:04 +08:00
@a.dlldefault = default alias i32, i32* @g.dlldefault
; CHECK: @a.dlldefault = alias i32, i32* @g.dlldefault
@a.dllexport = dllexport alias i32, i32* @g.dllexport
; CHECK: @a.dllexport = dllexport alias i32, i32* @g.dllexport
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
; Aliases -- ThreadLocal
[opaque pointer type] Add textual IR support for explicit type parameter for global aliases update.py: import fileinput import sys import re alias_match_prefix = r"(.*(?:=|:|^)\s*(?:external |)(?:(?:private|internal|linkonce|linkonce_odr|weak|weak_odr|common|appending|extern_weak|available_externally) )?(?:default |hidden |protected )?(?:dllimport |dllexport )?(?:unnamed_addr |)(?:thread_local(?:\([a-z]*\))? )?alias" plain = re.compile(alias_match_prefix + r" (.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|addrspacecast|\[\[[a-zA-Z]|\{\{).*$)") cast = re.compile(alias_match_prefix + r") ((?:bitcast|inttoptr|addrspacecast)\s*\(.* to (.*?)(| addrspace\(\d+\) *)\*\)\s*(?:;.*)?$)") gep = re.compile(alias_match_prefix + r") ((?:getelementptr)\s*(?:inbounds)?\s*\((?P<type>.*), (?P=type)(?:\s*addrspace\(\d+\)\s*)?\* .*\)\s*(?:;.*)?$)") def conv(line): m = re.match(cast, line) if m: return m.group(1) + " " + m.group(3) + ", " + m.group(2) m = re.match(gep, line) if m: return m.group(1) + " " + m.group(3) + ", " + m.group(2) m = re.match(plain, line) if m: return m.group(1) + ", " + m.group(2) + m.group(3) + "*" + m.group(4) + "\n" return line for line in sys.stdin: sys.stdout.write(conv(line)) apply.sh: for name in "$@" do python3 `dirname "$0"`/update.py < "$name" > "$name.tmp" && mv "$name.tmp" "$name" rm -f "$name.tmp" done The actual commands: From llvm/src: find test/ -name *.ll | xargs ./apply.sh From llvm/src/tools/clang: find test/ -name *.mm -o -name *.m -o -name *.cpp -o -name *.c | xargs -I '{}' ../../apply.sh "{}" From llvm/src/tools/polly: find test/ -name *.ll | xargs ./apply.sh llvm-svn: 247378
2015-09-11 11:22:04 +08:00
@a.notthreadlocal = alias i32, i32* @g.notthreadlocal
; CHECK: @a.notthreadlocal = alias i32, i32* @g.notthreadlocal
@a.generaldynamic = thread_local alias i32, i32* @g.generaldynamic
; CHECK: @a.generaldynamic = thread_local alias i32, i32* @g.generaldynamic
@a.localdynamic = thread_local(localdynamic) alias i32, i32* @g.localdynamic
; CHECK: @a.localdynamic = thread_local(localdynamic) alias i32, i32* @g.localdynamic
@a.initialexec = thread_local(initialexec) alias i32, i32* @g.initialexec
; CHECK: @a.initialexec = thread_local(initialexec) alias i32, i32* @g.initialexec
@a.localexec = thread_local(localexec) alias i32, i32* @g.localexec
; CHECK: @a.localexec = thread_local(localexec) alias i32, i32* @g.localexec
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
IR: Introduce local_unnamed_addr attribute. If a local_unnamed_addr attribute is attached to a global, the address is known to be insignificant within the module. It is distinct from the existing unnamed_addr attribute in that it only describes a local property of the module rather than a global property of the symbol. This attribute is intended to be used by the code generator and LTO to allow the linker to decide whether the global needs to be in the symbol table. It is possible to exclude a global from the symbol table if three things are true: - This attribute is present on every instance of the global (which means that the normal rule that the global must have a unique address can be broken without being observable by the program by performing comparisons against the global's address) - The global has linkonce_odr linkage (which means that each linkage unit must have its own copy of the global if it requires one, and the copy in each linkage unit must be the same) - It is a constant or a function (which means that the program cannot observe that the unique-address rule has been broken by writing to the global) Although this attribute could in principle be computed from the module contents, LTO clients (i.e. linkers) will normally need to be able to compute this property as part of symbol resolution, and it would be inefficient to materialize every module just to compute it. See: http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20160509/356401.html http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20160516/356738.html for earlier discussion. Part of the fix for PR27553. Differential Revision: http://reviews.llvm.org/D20348 llvm-svn: 272709
2016-06-15 05:01:22 +08:00
; Aliases -- unnamed_addr and local_unnamed_addr
[opaque pointer type] Add textual IR support for explicit type parameter for global aliases update.py: import fileinput import sys import re alias_match_prefix = r"(.*(?:=|:|^)\s*(?:external |)(?:(?:private|internal|linkonce|linkonce_odr|weak|weak_odr|common|appending|extern_weak|available_externally) )?(?:default |hidden |protected )?(?:dllimport |dllexport )?(?:unnamed_addr |)(?:thread_local(?:\([a-z]*\))? )?alias" plain = re.compile(alias_match_prefix + r" (.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|addrspacecast|\[\[[a-zA-Z]|\{\{).*$)") cast = re.compile(alias_match_prefix + r") ((?:bitcast|inttoptr|addrspacecast)\s*\(.* to (.*?)(| addrspace\(\d+\) *)\*\)\s*(?:;.*)?$)") gep = re.compile(alias_match_prefix + r") ((?:getelementptr)\s*(?:inbounds)?\s*\((?P<type>.*), (?P=type)(?:\s*addrspace\(\d+\)\s*)?\* .*\)\s*(?:;.*)?$)") def conv(line): m = re.match(cast, line) if m: return m.group(1) + " " + m.group(3) + ", " + m.group(2) m = re.match(gep, line) if m: return m.group(1) + " " + m.group(3) + ", " + m.group(2) m = re.match(plain, line) if m: return m.group(1) + ", " + m.group(2) + m.group(3) + "*" + m.group(4) + "\n" return line for line in sys.stdin: sys.stdout.write(conv(line)) apply.sh: for name in "$@" do python3 `dirname "$0"`/update.py < "$name" > "$name.tmp" && mv "$name.tmp" "$name" rm -f "$name.tmp" done The actual commands: From llvm/src: find test/ -name *.ll | xargs ./apply.sh From llvm/src/tools/clang: find test/ -name *.mm -o -name *.m -o -name *.cpp -o -name *.c | xargs -I '{}' ../../apply.sh "{}" From llvm/src/tools/polly: find test/ -name *.ll | xargs ./apply.sh llvm-svn: 247378
2015-09-11 11:22:04 +08:00
@a.unnamed_addr = unnamed_addr alias i32, i32* @g.unnamed_addr
; CHECK: @a.unnamed_addr = unnamed_addr alias i32, i32* @g.unnamed_addr
IR: Introduce local_unnamed_addr attribute. If a local_unnamed_addr attribute is attached to a global, the address is known to be insignificant within the module. It is distinct from the existing unnamed_addr attribute in that it only describes a local property of the module rather than a global property of the symbol. This attribute is intended to be used by the code generator and LTO to allow the linker to decide whether the global needs to be in the symbol table. It is possible to exclude a global from the symbol table if three things are true: - This attribute is present on every instance of the global (which means that the normal rule that the global must have a unique address can be broken without being observable by the program by performing comparisons against the global's address) - The global has linkonce_odr linkage (which means that each linkage unit must have its own copy of the global if it requires one, and the copy in each linkage unit must be the same) - It is a constant or a function (which means that the program cannot observe that the unique-address rule has been broken by writing to the global) Although this attribute could in principle be computed from the module contents, LTO clients (i.e. linkers) will normally need to be able to compute this property as part of symbol resolution, and it would be inefficient to materialize every module just to compute it. See: http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20160509/356401.html http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20160516/356738.html for earlier discussion. Part of the fix for PR27553. Differential Revision: http://reviews.llvm.org/D20348 llvm-svn: 272709
2016-06-15 05:01:22 +08:00
@a.local_unnamed_addr = local_unnamed_addr alias i32, i32* @g.local_unnamed_addr
; CHECK: @a.local_unnamed_addr = local_unnamed_addr alias i32, i32* @g.local_unnamed_addr
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
Add IR support, ELF section and user documentation for partitioning feature. The partitioning feature was proposed here: http://lists.llvm.org/pipermail/llvm-dev/2019-February/130583.html This is mostly just documentation. The feature itself will be contributed in subsequent patches. Differential Revision: https://reviews.llvm.org/D60242 llvm-svn: 361923
2019-05-29 11:29:01 +08:00
; Aliases -- partition
; CHECK: @alias.partition = alias i32, i32* @g.partition, partition "part"
@alias.partition = alias i32, i32* @g.partition, partition "part"
[GCC] Attribute ifunc support in llvm This patch add support for GCC attribute((ifunc("resolver"))) for targets that use ELF as object file format. In general ifunc is a special kind of function alias with type @gnu_indirect_function. Patch for Clang http://reviews.llvm.org/D15524 Differential Revision: http://reviews.llvm.org/D15525 llvm-svn: 265667
2016-04-07 20:32:19 +08:00
;; IFunc
; Format @<Name> = [Linkage] [Visibility] ifunc <IFuncTy>,
; <ResolverTy>* @<Resolver>
; IFunc -- Linkage
@ifunc.external = external ifunc void (), i8* ()* @ifunc_resolver
; CHECK: @ifunc.external = ifunc void (), i8* ()* @ifunc_resolver
@ifunc.private = private ifunc void (), i8* ()* @ifunc_resolver
; CHECK: @ifunc.private = private ifunc void (), i8* ()* @ifunc_resolver
@ifunc.internal = internal ifunc void (), i8* ()* @ifunc_resolver
; CHECK: @ifunc.internal = internal ifunc void (), i8* ()* @ifunc_resolver
; IFunc -- Visibility
@ifunc.default = default ifunc void (), i8* ()* @ifunc_resolver
; CHECK: @ifunc.default = ifunc void (), i8* ()* @ifunc_resolver
@ifunc.hidden = hidden ifunc void (), i8* ()* @ifunc_resolver
; CHECK: @ifunc.hidden = hidden ifunc void (), i8* ()* @ifunc_resolver
@ifunc.protected = protected ifunc void (), i8* ()* @ifunc_resolver
; CHECK: @ifunc.protected = protected ifunc void (), i8* ()* @ifunc_resolver
Add IR support, ELF section and user documentation for partitioning feature. The partitioning feature was proposed here: http://lists.llvm.org/pipermail/llvm-dev/2019-February/130583.html This is mostly just documentation. The feature itself will be contributed in subsequent patches. Differential Revision: https://reviews.llvm.org/D60242 llvm-svn: 361923
2019-05-29 11:29:01 +08:00
; IFunc -- partition
; CHECK: @ifunc.partition = ifunc void (), i8* ()* @ifunc_resolver, partition "part"
@ifunc.partition = ifunc void (), i8* ()* @ifunc_resolver, partition "part"
[GCC] Attribute ifunc support in llvm This patch add support for GCC attribute((ifunc("resolver"))) for targets that use ELF as object file format. In general ifunc is a special kind of function alias with type @gnu_indirect_function. Patch for Clang http://reviews.llvm.org/D15524 Differential Revision: http://reviews.llvm.org/D15525 llvm-svn: 265667
2016-04-07 20:32:19 +08:00
define i8* @ifunc_resolver() {
entry:
ret i8* null
}
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
;; Functions
; Format: define [linkage] [visibility] [DLLStorageClass]
; [cconv] [ret attrs]
; <ResultType> @<FunctionName> ([argument list])
IR: Introduce local_unnamed_addr attribute. If a local_unnamed_addr attribute is attached to a global, the address is known to be insignificant within the module. It is distinct from the existing unnamed_addr attribute in that it only describes a local property of the module rather than a global property of the symbol. This attribute is intended to be used by the code generator and LTO to allow the linker to decide whether the global needs to be in the symbol table. It is possible to exclude a global from the symbol table if three things are true: - This attribute is present on every instance of the global (which means that the normal rule that the global must have a unique address can be broken without being observable by the program by performing comparisons against the global's address) - The global has linkonce_odr linkage (which means that each linkage unit must have its own copy of the global if it requires one, and the copy in each linkage unit must be the same) - It is a constant or a function (which means that the program cannot observe that the unique-address rule has been broken by writing to the global) Although this attribute could in principle be computed from the module contents, LTO clients (i.e. linkers) will normally need to be able to compute this property as part of symbol resolution, and it would be inefficient to materialize every module just to compute it. See: http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20160509/356401.html http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20160516/356738.html for earlier discussion. Part of the fix for PR27553. Differential Revision: http://reviews.llvm.org/D20348 llvm-svn: 272709
2016-06-15 05:01:22 +08:00
; [(unnamed_addr|local_unnamed_addr)] [fn Attrs] [section "name"] [comdat [($name)]]
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
; [align N] [gc] [prefix Constant] [prologue Constant]
; [personality Constant] { ... }
; Functions -- Simple
declare void @f1 ()
; CHECK: declare void @f1()
define void @f2 () {
; CHECK: define void @f2()
entry:
ret void
}
; Functions -- linkage
define private void @f.private() {
; CHECK: define private void @f.private()
entry:
ret void
}
define internal void @f.internal() {
; CHECK: define internal void @f.internal()
entry:
ret void
}
define available_externally void @f.available_externally() {
; CHECK: define available_externally void @f.available_externally()
entry:
ret void
}
define linkonce void @f.linkonce() {
; CHECK: define linkonce void @f.linkonce()
entry:
ret void
}
define weak void @f.weak() {
; CHECK: define weak void @f.weak()
entry:
ret void
}
define linkonce_odr void @f.linkonce_odr() {
; CHECK: define linkonce_odr void @f.linkonce_odr()
entry:
ret void
}
define weak_odr void @f.weak_odr() {
; CHECK: define weak_odr void @f.weak_odr()
entry:
ret void
}
declare external void @f.external()
; CHECK: declare void @f.external()
declare extern_weak void @f.extern_weak()
; CHECK: declare extern_weak void @f.extern_weak()
; Functions -- visibility
declare default void @f.default()
; CHECK: declare void @f.default()
declare hidden void @f.hidden()
; CHECK: declare hidden void @f.hidden()
declare protected void @f.protected()
; CHECK: declare protected void @f.protected()
; Functions -- DLLStorageClass
declare dllimport void @f.dllimport()
; CHECK: declare dllimport void @f.dllimport()
declare dllexport void @f.dllexport()
; CHECK: declare dllexport void @f.dllexport()
; Functions -- cconv (Calling conventions)
declare ccc void @f.ccc()
; CHECK: declare void @f.ccc()
declare fastcc void @f.fastcc()
; CHECK: declare fastcc void @f.fastcc()
declare coldcc void @f.coldcc()
; CHECK: declare coldcc void @f.coldcc()
declare cc10 void @f.cc10()
; CHECK: declare ghccc void @f.cc10()
declare ghccc void @f.ghccc()
; CHECK: declare ghccc void @f.ghccc()
declare cc11 void @f.cc11()
; CHECK: declare cc11 void @f.cc11()
declare webkit_jscc void @f.webkit_jscc()
; CHECK: declare webkit_jscc void @f.webkit_jscc()
declare anyregcc void @f.anyregcc()
; CHECK: declare anyregcc void @f.anyregcc()
declare preserve_mostcc void @f.preserve_mostcc()
; CHECK: declare preserve_mostcc void @f.preserve_mostcc()
declare preserve_allcc void @f.preserve_allcc()
; CHECK: declare preserve_allcc void @f.preserve_allcc()
declare cc64 void @f.cc64()
; CHECK: declare x86_stdcallcc void @f.cc64()
declare x86_stdcallcc void @f.x86_stdcallcc()
; CHECK: declare x86_stdcallcc void @f.x86_stdcallcc()
declare cc65 void @f.cc65()
; CHECK: declare x86_fastcallcc void @f.cc65()
declare x86_fastcallcc void @f.x86_fastcallcc()
; CHECK: declare x86_fastcallcc void @f.x86_fastcallcc()
declare cc66 void @f.cc66()
; CHECK: declare arm_apcscc void @f.cc66()
declare arm_apcscc void @f.arm_apcscc()
; CHECK: declare arm_apcscc void @f.arm_apcscc()
declare cc67 void @f.cc67()
; CHECK: declare arm_aapcscc void @f.cc67()
declare arm_aapcscc void @f.arm_aapcscc()
; CHECK: declare arm_aapcscc void @f.arm_aapcscc()
declare cc68 void @f.cc68()
; CHECK: declare arm_aapcs_vfpcc void @f.cc68()
declare arm_aapcs_vfpcc void @f.arm_aapcs_vfpcc()
; CHECK: declare arm_aapcs_vfpcc void @f.arm_aapcs_vfpcc()
declare cc69 void @f.cc69()
; CHECK: declare msp430_intrcc void @f.cc69()
declare msp430_intrcc void @f.msp430_intrcc()
; CHECK: declare msp430_intrcc void @f.msp430_intrcc()
declare cc70 void @f.cc70()
; CHECK: declare x86_thiscallcc void @f.cc70()
declare x86_thiscallcc void @f.x86_thiscallcc()
; CHECK: declare x86_thiscallcc void @f.x86_thiscallcc()
declare cc71 void @f.cc71()
; CHECK: declare ptx_kernel void @f.cc71()
declare ptx_kernel void @f.ptx_kernel()
; CHECK: declare ptx_kernel void @f.ptx_kernel()
declare cc72 void @f.cc72()
; CHECK: declare ptx_device void @f.cc72()
declare ptx_device void @f.ptx_device()
; CHECK: declare ptx_device void @f.ptx_device()
declare cc75 void @f.cc75()
; CHECK: declare spir_func void @f.cc75()
declare spir_func void @f.spir_func()
; CHECK: declare spir_func void @f.spir_func()
declare cc76 void @f.cc76()
; CHECK: declare spir_kernel void @f.cc76()
declare spir_kernel void @f.spir_kernel()
; CHECK: declare spir_kernel void @f.spir_kernel()
declare cc77 void @f.cc77()
; CHECK: declare intel_ocl_bicc void @f.cc77()
declare intel_ocl_bicc void @f.intel_ocl_bicc()
; CHECK: declare intel_ocl_bicc void @f.intel_ocl_bicc()
declare cc78 void @f.cc78()
; CHECK: declare x86_64_sysvcc void @f.cc78()
declare x86_64_sysvcc void @f.x86_64_sysvcc()
; CHECK: declare x86_64_sysvcc void @f.x86_64_sysvcc()
declare cc79 void @f.cc79()
[AArch64] Extend CallingConv::X86_64_Win64 to AArch64 as well Rename the enum value from X86_64_Win64 to plain Win64. The symbol exposed in the textual IR is changed from 'x86_64_win64cc' to 'win64cc', but the numeric value is kept, keeping support for old bitcode. Differential Revision: https://reviews.llvm.org/D34474 llvm-svn: 308208
2017-07-18 04:05:19 +08:00
; CHECK: declare win64cc void @f.cc79()
declare win64cc void @f.win64cc()
; CHECK: declare win64cc void @f.win64cc()
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
declare cc80 void @f.cc80()
; CHECK: declare x86_vectorcallcc void @f.cc80()
declare x86_vectorcallcc void @f.x86_vectorcallcc()
; CHECK: declare x86_vectorcallcc void @f.x86_vectorcallcc()
AMDGPU/SI: Add amdgpu_kernel calling convention. Part 1. Summary: This will be used for AMDGPU_HSA_KERNEL symbol type in output ELF. Also, in the future unused non-kernels may be optimized. For now, also accept SPIR_KERNEL for HCC frontend. Also, add bitcode compatibility tests for missing calling conventions except AVR_BUILTIN which doesn't have parse code. Reviewers: tstellarAMD, arsenm Subscribers: arsenm, joker.eph, llvm-commits llvm-svn: 268717
2016-05-06 17:07:29 +08:00
declare cc81 void @f.cc81()
; CHECK: declare hhvmcc void @f.cc81()
declare hhvmcc void @f.hhvmcc()
; CHECK: declare hhvmcc void @f.hhvmcc()
declare cc82 void @f.cc82()
; CHECK: declare hhvm_ccc void @f.cc82()
declare hhvm_ccc void @f.hhvm_ccc()
; CHECK: declare hhvm_ccc void @f.hhvm_ccc()
declare cc83 void @f.cc83()
; CHECK: declare x86_intrcc void @f.cc83()
declare x86_intrcc void @f.x86_intrcc()
; CHECK: declare x86_intrcc void @f.x86_intrcc()
declare cc84 void @f.cc84()
; CHECK: declare avr_intrcc void @f.cc84()
declare avr_intrcc void @f.avr_intrcc()
; CHECK: declare avr_intrcc void @f.avr_intrcc()
declare cc85 void @f.cc85()
; CHECK: declare avr_signalcc void @f.cc85()
declare avr_signalcc void @f.avr_signalcc()
; CHECK: declare avr_signalcc void @f.avr_signalcc()
declare cc87 void @f.cc87()
; CHECK: declare amdgpu_vs void @f.cc87()
declare amdgpu_vs void @f.amdgpu_vs()
; CHECK: declare amdgpu_vs void @f.amdgpu_vs()
declare cc88 void @f.cc88()
; CHECK: declare amdgpu_gs void @f.cc88()
declare amdgpu_gs void @f.amdgpu_gs()
; CHECK: declare amdgpu_gs void @f.amdgpu_gs()
declare cc89 void @f.cc89()
; CHECK: declare amdgpu_ps void @f.cc89()
declare amdgpu_ps void @f.amdgpu_ps()
; CHECK: declare amdgpu_ps void @f.amdgpu_ps()
declare cc90 void @f.cc90()
; CHECK: declare amdgpu_cs void @f.cc90()
declare amdgpu_cs void @f.amdgpu_cs()
; CHECK: declare amdgpu_cs void @f.amdgpu_cs()
declare cc91 void @f.cc91()
; CHECK: declare amdgpu_kernel void @f.cc91()
declare amdgpu_kernel void @f.amdgpu_kernel()
; CHECK: declare amdgpu_kernel void @f.amdgpu_kernel()
AMDGPU: Add AMDGPU_HS calling convention Reviewers: arsenm, nhaehnle Subscribers: mehdi_amini, kzhuravl, wdng, yaxunl, dstuttard, tpr, llvm-commits, t-tye Differential Revision: https://reviews.llvm.org/D32644 llvm-svn: 301930
2017-05-02 23:41:10 +08:00
declare cc93 void @f.cc93()
; CHECK: declare amdgpu_hs void @f.cc93()
declare amdgpu_hs void @f.amdgpu_hs()
; CHECK: declare amdgpu_hs void @f.amdgpu_hs()
[AMDGPU] calling conventions for AMDPAL OS type Summary: This commit adds comments on how the AMDPAL OS type overloads the existing AMDGPU_ calling conventions used by Mesa, and adds a couple of new ones. Reviewers: arsenm, nhaehnle, dstuttard Subscribers: mehdi_amini, kzhuravl, wdng, yaxunl, t-tye, llvm-commits Differential Revision: https://reviews.llvm.org/D37752 llvm-svn: 314502
2017-09-29 17:51:22 +08:00
declare cc95 void @f.cc95()
; CHECK: declare amdgpu_ls void @f.cc95()
declare amdgpu_ls void @f.amdgpu_ls()
; CHECK: declare amdgpu_ls void @f.amdgpu_ls()
declare cc96 void @f.cc96()
; CHECK: declare amdgpu_es void @f.cc96()
declare amdgpu_es void @f.amdgpu_es()
; CHECK: declare amdgpu_es void @f.amdgpu_es()
[IR] Limit bits used for CallingConv::ID, update tests Use 10 bits to represent calling convention ID's instead of 13, and update the bitcode compatibility tests accordingly. We now error-out in the bitcode reader when we see bad calling conv ID's. Thanks to rnk and dexonsmith for feedback! Differential Revision: http://reviews.llvm.org/D13826 llvm-svn: 251452
2015-10-28 05:17:06 +08:00
declare cc1023 void @f.cc1023()
; CHECK: declare cc1023 void @f.cc1023()
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
; Functions -- ret attrs (Return attributes)
declare zeroext i64 @f.zeroext()
; CHECK: declare zeroext i64 @f.zeroext()
declare signext i64 @f.signext()
; CHECK: declare signext i64 @f.signext()
declare inreg i32* @f.inreg()
; CHECK: declare inreg i32* @f.inreg()
declare noalias i32* @f.noalias()
; CHECK: declare noalias i32* @f.noalias()
declare nonnull i32* @f.nonnull()
; CHECK: declare nonnull i32* @f.nonnull()
declare dereferenceable(4) i32* @f.dereferenceable4()
; CHECK: declare dereferenceable(4) i32* @f.dereferenceable4()
declare dereferenceable(8) i32* @f.dereferenceable8()
; CHECK: declare dereferenceable(8) i32* @f.dereferenceable8()
declare dereferenceable(16) i32* @f.dereferenceable16()
; CHECK: declare dereferenceable(16) i32* @f.dereferenceable16()
declare dereferenceable_or_null(4) i32* @f.dereferenceable4_or_null()
; CHECK: declare dereferenceable_or_null(4) i32* @f.dereferenceable4_or_null()
declare dereferenceable_or_null(8) i32* @f.dereferenceable8_or_null()
; CHECK: declare dereferenceable_or_null(8) i32* @f.dereferenceable8_or_null()
declare dereferenceable_or_null(16) i32* @f.dereferenceable16_or_null()
; CHECK: declare dereferenceable_or_null(16) i32* @f.dereferenceable16_or_null()
; Functions -- Parameter attributes
declare void @f.param.zeroext(i8 zeroext)
; CHECK: declare void @f.param.zeroext(i8 zeroext)
declare void @f.param.signext(i8 signext)
; CHECK: declare void @f.param.signext(i8 signext)
declare void @f.param.inreg(i8 inreg)
; CHECK: declare void @f.param.inreg(i8 inreg)
declare void @f.param.byval({ i8, i8 }* byval)
Reapply: IR: add optional type to 'byval' function parameters When we switch to opaque pointer types we will need some way to describe how many bytes a 'byval' parameter should occupy on the stack. This adds a (for now) optional extra type parameter. If present, the type must match the pointee type of the argument. The original commit did not remap byval types when linking modules, which broke LTO. This version fixes that. Note to front-end maintainers: if this causes test failures, it's probably because the "byval" attribute is printed after attributes without any parameter after this change. llvm-svn: 362128
2019-05-31 02:48:23 +08:00
; CHECK: declare void @f.param.byval({ i8, i8 }* byval({ i8, i8 }))
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
declare void @f.param.inalloca(i8* inalloca)
; CHECK: declare void @f.param.inalloca(i8* inalloca)
declare void @f.param.sret(i8* sret)
; CHECK: declare void @f.param.sret(i8* sret)
declare void @f.param.noalias(i8* noalias)
; CHECK: declare void @f.param.noalias(i8* noalias)
declare void @f.param.nocapture(i8* nocapture)
; CHECK: declare void @f.param.nocapture(i8* nocapture)
declare void @f.param.nest(i8* nest)
; CHECK: declare void @f.param.nest(i8* nest)
declare i8* @f.param.returned(i8* returned)
; CHECK: declare i8* @f.param.returned(i8* returned)
declare void @f.param.nonnull(i8* nonnull)
; CHECK: declare void @f.param.nonnull(i8* nonnull)
declare void @f.param.dereferenceable(i8* dereferenceable(4))
; CHECK: declare void @f.param.dereferenceable(i8* dereferenceable(4))
declare void @f.param.dereferenceable_or_null(i8* dereferenceable_or_null(4))
; CHECK: declare void @f.param.dereferenceable_or_null(i8* dereferenceable_or_null(4))
IR: Introduce local_unnamed_addr attribute. If a local_unnamed_addr attribute is attached to a global, the address is known to be insignificant within the module. It is distinct from the existing unnamed_addr attribute in that it only describes a local property of the module rather than a global property of the symbol. This attribute is intended to be used by the code generator and LTO to allow the linker to decide whether the global needs to be in the symbol table. It is possible to exclude a global from the symbol table if three things are true: - This attribute is present on every instance of the global (which means that the normal rule that the global must have a unique address can be broken without being observable by the program by performing comparisons against the global's address) - The global has linkonce_odr linkage (which means that each linkage unit must have its own copy of the global if it requires one, and the copy in each linkage unit must be the same) - It is a constant or a function (which means that the program cannot observe that the unique-address rule has been broken by writing to the global) Although this attribute could in principle be computed from the module contents, LTO clients (i.e. linkers) will normally need to be able to compute this property as part of symbol resolution, and it would be inefficient to materialize every module just to compute it. See: http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20160509/356401.html http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20160516/356738.html for earlier discussion. Part of the fix for PR27553. Differential Revision: http://reviews.llvm.org/D20348 llvm-svn: 272709
2016-06-15 05:01:22 +08:00
; Functions -- unnamed_addr and local_unnamed_addr
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
declare void @f.unnamed_addr() unnamed_addr
; CHECK: declare void @f.unnamed_addr() unnamed_addr
IR: Introduce local_unnamed_addr attribute. If a local_unnamed_addr attribute is attached to a global, the address is known to be insignificant within the module. It is distinct from the existing unnamed_addr attribute in that it only describes a local property of the module rather than a global property of the symbol. This attribute is intended to be used by the code generator and LTO to allow the linker to decide whether the global needs to be in the symbol table. It is possible to exclude a global from the symbol table if three things are true: - This attribute is present on every instance of the global (which means that the normal rule that the global must have a unique address can be broken without being observable by the program by performing comparisons against the global's address) - The global has linkonce_odr linkage (which means that each linkage unit must have its own copy of the global if it requires one, and the copy in each linkage unit must be the same) - It is a constant or a function (which means that the program cannot observe that the unique-address rule has been broken by writing to the global) Although this attribute could in principle be computed from the module contents, LTO clients (i.e. linkers) will normally need to be able to compute this property as part of symbol resolution, and it would be inefficient to materialize every module just to compute it. See: http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20160509/356401.html http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20160516/356738.html for earlier discussion. Part of the fix for PR27553. Differential Revision: http://reviews.llvm.org/D20348 llvm-svn: 272709
2016-06-15 05:01:22 +08:00
declare void @f.local_unnamed_addr() local_unnamed_addr
; CHECK: declare void @f.local_unnamed_addr() local_unnamed_addr
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
; Functions -- fn Attrs (Function attributes)
declare void @f.alignstack4() alignstack(4)
; CHECK: declare void @f.alignstack4() #0
declare void @f.alignstack8() alignstack(8)
; CHECK: declare void @f.alignstack8() #1
declare void @f.alwaysinline() alwaysinline
; CHECK: declare void @f.alwaysinline() #2
declare void @f.cold() cold
; CHECK: declare void @f.cold() #3
declare void @f.convergent() convergent
; CHECK: declare void @f.convergent() #4
declare void @f.inlinehint() inlinehint
; CHECK: declare void @f.inlinehint() #5
declare void @f.jumptable() unnamed_addr jumptable
; CHECK: declare void @f.jumptable() unnamed_addr #6
declare void @f.minsize() minsize
; CHECK: declare void @f.minsize() #7
declare void @f.naked() naked
; CHECK: declare void @f.naked() #8
declare void @f.nobuiltin() nobuiltin
; CHECK: declare void @f.nobuiltin() #9
declare void @f.noduplicate() noduplicate
; CHECK: declare void @f.noduplicate() #10
declare void @f.noimplicitfloat() noimplicitfloat
; CHECK: declare void @f.noimplicitfloat() #11
declare void @f.noinline() noinline
; CHECK: declare void @f.noinline() #12
declare void @f.nonlazybind() nonlazybind
; CHECK: declare void @f.nonlazybind() #13
declare void @f.noredzone() noredzone
; CHECK: declare void @f.noredzone() #14
declare void @f.noreturn() noreturn
; CHECK: declare void @f.noreturn() #15
declare void @f.nounwind() nounwind
; CHECK: declare void @f.nounwind() #16
declare void @f.optnone() noinline optnone
; CHECK: declare void @f.optnone() #17
declare void @f.optsize() optsize
; CHECK: declare void @f.optsize() #18
declare void @f.readnone() readnone
; CHECK: declare void @f.readnone() #19
declare void @f.readonly() readonly
; CHECK: declare void @f.readonly() #20
declare void @f.returns_twice() returns_twice
; CHECK: declare void @f.returns_twice() #21
declare void @f.safestack() safestack
; CHECK: declare void @f.safestack() #22
declare void @f.sanitize_address() sanitize_address
; CHECK: declare void @f.sanitize_address() #23
declare void @f.sanitize_memory() sanitize_memory
; CHECK: declare void @f.sanitize_memory() #24
declare void @f.sanitize_thread() sanitize_thread
; CHECK: declare void @f.sanitize_thread() #25
declare void @f.ssp() ssp
; CHECK: declare void @f.ssp() #26
declare void @f.sspreq() sspreq
; CHECK: declare void @f.sspreq() #27
declare void @f.sspstrong() sspstrong
; CHECK: declare void @f.sspstrong() #28
declare void @f.thunk() "thunk"
; CHECK: declare void @f.thunk() #29
declare void @f.uwtable() uwtable
; CHECK: declare void @f.uwtable() #30
declare void @f.kvpair() "cpu"="cortex-a8"
; CHECK:declare void @f.kvpair() #31
Add a new attribute: norecurse This attribute allows the compiler to assume that the function never recurses into itself, either directly or indirectly (transitively). This can be used among other things to demote global variables to locals. llvm-svn: 252282
2015-11-06 18:32:53 +08:00
declare void @f.norecurse() norecurse
; CHECK: declare void @f.norecurse() #32
Add InaccessibleMemOnly and inaccessibleMemOrArgMemOnly attributes Summary: This patch introduces two new function attributes InaccessibleMemOnly: This attribute indicates that the function may only access memory that is not accessible by the program/IR being compiled. This is a weaker form of ReadNone. inaccessibleMemOrArgMemOnly: This attribute indicates that the function may only access memory that is either not accessible by the program/IR being compiled, or is pointed to by its pointer arguments. This is a weaker form of ArgMemOnly Test cases have been updated. This revision uses this (https://github.com/llvm-mirror/llvm/commit/d001932f3a8aa1ebd1555162fdce365f011bc292) as reference. Reviewers: jmolloy, hfinkel Subscribers: reames, joker.eph, llvm-commits Differential Revision: http://reviews.llvm.org/D15499 llvm-svn: 255778
2015-12-17 00:16:19 +08:00
declare void @f.inaccessiblememonly() inaccessiblememonly
; CHECK: declare void @f.inaccessiblememonly() #33
declare void @f.inaccessiblemem_or_argmemonly() inaccessiblemem_or_argmemonly
; CHECK: declare void @f.inaccessiblemem_or_argmemonly() #34
Add strictfp attribute to prevent unwanted optimizations of libm calls Differential Revision: https://reviews.llvm.org/D34163 llvm-svn: 310885
2017-08-15 05:15:13 +08:00
declare void @f.strictfp() #35
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
; Functions -- section
declare void @f.section() section "80"
; CHECK: declare void @f.section() section "80"
Add IR support, ELF section and user documentation for partitioning feature. The partitioning feature was proposed here: http://lists.llvm.org/pipermail/llvm-dev/2019-February/130583.html This is mostly just documentation. The feature itself will be contributed in subsequent patches. Differential Revision: https://reviews.llvm.org/D60242 llvm-svn: 361923
2019-05-29 11:29:01 +08:00
; Functions -- partition
define void @f.partition() partition "part" {
; CHECK: define void @f.partition() partition "part"
ret void
}
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
; Functions -- comdat
define void @f.comdat_any() comdat($comdat.any) {
; CHECK: define void @f.comdat_any() comdat($comdat.any)
entry:
ret void
}
define void @f.comdat_exactmatch() comdat($comdat.exactmatch) {
; CHECK: define void @f.comdat_exactmatch() comdat($comdat.exactmatch)
entry:
ret void
}
define void @f.comdat_largest() comdat($comdat.largest) {
; CHECK: define void @f.comdat_largest() comdat($comdat.largest)
entry:
ret void
}
define void @f.comdat_noduplicates() comdat($comdat.noduplicates) {
; CHECK: define void @f.comdat_noduplicates() comdat($comdat.noduplicates)
entry:
ret void
}
define void @f.comdat_samesize() comdat($comdat.samesize) {
; CHECK: define void @f.comdat_samesize() comdat($comdat.samesize)
entry:
ret void
}
; Functions -- align
declare void @f.align2() align 2
; CHECK: declare void @f.align2() align 2
declare void @f.align4() align 4
; CHECK: declare void @f.align4() align 4
declare void @f.align8() align 8
; CHECK: declare void @f.align8() align 8
; Functions -- GC
declare void @f.gcshadow() gc "shadow-stack"
; CHECK: declare void @f.gcshadow() gc "shadow-stack"
; Functions -- Prefix data
declare void @f.prefixi32() prefix i32 1684365668
; CHECK: declare void @f.prefixi32() prefix i32 1684365668
declare void @f.prefixarray() prefix [4 x i32] [i32 0, i32 1, i32 2, i32 3]
; CHECK: declare void @f.prefixarray() prefix [4 x i32] [i32 0, i32 1, i32 2, i32 3]
; Functions -- Prologue data
declare void @f.prologuei32() prologue i32 1684365669
; CHECK: declare void @f.prologuei32() prologue i32 1684365669
declare void @f.prologuearray() prologue [4 x i32] [i32 0, i32 1, i32 2, i32 3]
; CHECK: declare void @f.prologuearray() prologue [4 x i32] [i32 0, i32 1, i32 2, i32 3]
; Functions -- Personality constant
declare void @llvm.donothing() nounwind readnone
Add InaccessibleMemOnly and inaccessibleMemOrArgMemOnly attributes Summary: This patch introduces two new function attributes InaccessibleMemOnly: This attribute indicates that the function may only access memory that is not accessible by the program/IR being compiled. This is a weaker form of ReadNone. inaccessibleMemOrArgMemOnly: This attribute indicates that the function may only access memory that is either not accessible by the program/IR being compiled, or is pointed to by its pointer arguments. This is a weaker form of ArgMemOnly Test cases have been updated. This revision uses this (https://github.com/llvm-mirror/llvm/commit/d001932f3a8aa1ebd1555162fdce365f011bc292) as reference. Reviewers: jmolloy, hfinkel Subscribers: reames, joker.eph, llvm-commits Differential Revision: http://reviews.llvm.org/D15499 llvm-svn: 255778
2015-12-17 00:16:19 +08:00
; CHECK: declare void @llvm.donothing() #35
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
define void @f.no_personality() personality i8 3 {
; CHECK: define void @f.no_personality() personality i8 3
invoke void @llvm.donothing() to label %normal unwind label %exception
exception:
%cleanup = landingpad i8 cleanup
br label %normal
normal:
ret void
}
declare i32 @f.personality_handler()
; CHECK: declare i32 @f.personality_handler()
define void @f.personality() personality i32 ()* @f.personality_handler {
; CHECK: define void @f.personality() personality i32 ()* @f.personality_handler
invoke void @llvm.donothing() to label %normal unwind label %exception
exception:
%cleanup = landingpad i32 cleanup
br label %normal
normal:
ret void
}
;; Atomic Memory Ordering Constraints
define void @atomics(i32* %word) {
%cmpxchg.0 = cmpxchg i32* %word, i32 0, i32 4 monotonic monotonic
; CHECK: %cmpxchg.0 = cmpxchg i32* %word, i32 0, i32 4 monotonic monotonic
%cmpxchg.1 = cmpxchg i32* %word, i32 0, i32 5 acq_rel monotonic
; CHECK: %cmpxchg.1 = cmpxchg i32* %word, i32 0, i32 5 acq_rel monotonic
%cmpxchg.2 = cmpxchg i32* %word, i32 0, i32 6 acquire monotonic
; CHECK: %cmpxchg.2 = cmpxchg i32* %word, i32 0, i32 6 acquire monotonic
%cmpxchg.3 = cmpxchg i32* %word, i32 0, i32 7 release monotonic
; CHECK: %cmpxchg.3 = cmpxchg i32* %word, i32 0, i32 7 release monotonic
%cmpxchg.4 = cmpxchg i32* %word, i32 0, i32 8 seq_cst monotonic
; CHECK: %cmpxchg.4 = cmpxchg i32* %word, i32 0, i32 8 seq_cst monotonic
%cmpxchg.5 = cmpxchg weak i32* %word, i32 0, i32 9 seq_cst monotonic
; CHECK: %cmpxchg.5 = cmpxchg weak i32* %word, i32 0, i32 9 seq_cst monotonic
%cmpxchg.6 = cmpxchg volatile i32* %word, i32 0, i32 10 seq_cst monotonic
; CHECK: %cmpxchg.6 = cmpxchg volatile i32* %word, i32 0, i32 10 seq_cst monotonic
Enhance synchscope representation OpenCL 2.0 introduces the notion of memory scopes in atomic operations to global and local memory. These scopes restrict how synchronization is achieved, which can result in improved performance. This change extends existing notion of synchronization scopes in LLVM to support arbitrary scopes expressed as target-specific strings, in addition to the already defined scopes (single thread, system). The LLVM IR and MIR syntax for expressing synchronization scopes has changed to use *syncscope("<scope>")*, where <scope> can be "singlethread" (this replaces *singlethread* keyword), or a target-specific name. As before, if the scope is not specified, it defaults to CrossThread/System scope. Implementation details: - Mapping from synchronization scope name/string to synchronization scope id is stored in LLVM context; - CrossThread/System and SingleThread scopes are pre-defined to efficiently check for known scopes without comparing strings; - Synchronization scope names are stored in SYNC_SCOPE_NAMES_BLOCK in the bitcode. Differential Revision: https://reviews.llvm.org/D21723 llvm-svn: 307722
2017-07-12 06:23:00 +08:00
%cmpxchg.7 = cmpxchg weak volatile i32* %word, i32 0, i32 11 syncscope("singlethread") seq_cst monotonic
; CHECK: %cmpxchg.7 = cmpxchg weak volatile i32* %word, i32 0, i32 11 syncscope("singlethread") seq_cst monotonic
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
%atomicrmw.xchg = atomicrmw xchg i32* %word, i32 12 monotonic
; CHECK: %atomicrmw.xchg = atomicrmw xchg i32* %word, i32 12 monotonic
%atomicrmw.add = atomicrmw add i32* %word, i32 13 monotonic
; CHECK: %atomicrmw.add = atomicrmw add i32* %word, i32 13 monotonic
%atomicrmw.sub = atomicrmw sub i32* %word, i32 14 monotonic
; CHECK: %atomicrmw.sub = atomicrmw sub i32* %word, i32 14 monotonic
%atomicrmw.and = atomicrmw and i32* %word, i32 15 monotonic
; CHECK: %atomicrmw.and = atomicrmw and i32* %word, i32 15 monotonic
%atomicrmw.nand = atomicrmw nand i32* %word, i32 16 monotonic
; CHECK: %atomicrmw.nand = atomicrmw nand i32* %word, i32 16 monotonic
%atomicrmw.or = atomicrmw or i32* %word, i32 17 monotonic
; CHECK: %atomicrmw.or = atomicrmw or i32* %word, i32 17 monotonic
%atomicrmw.xor = atomicrmw xor i32* %word, i32 18 monotonic
; CHECK: %atomicrmw.xor = atomicrmw xor i32* %word, i32 18 monotonic
%atomicrmw.max = atomicrmw max i32* %word, i32 19 monotonic
; CHECK: %atomicrmw.max = atomicrmw max i32* %word, i32 19 monotonic
%atomicrmw.min = atomicrmw volatile min i32* %word, i32 20 monotonic
; CHECK: %atomicrmw.min = atomicrmw volatile min i32* %word, i32 20 monotonic
Enhance synchscope representation OpenCL 2.0 introduces the notion of memory scopes in atomic operations to global and local memory. These scopes restrict how synchronization is achieved, which can result in improved performance. This change extends existing notion of synchronization scopes in LLVM to support arbitrary scopes expressed as target-specific strings, in addition to the already defined scopes (single thread, system). The LLVM IR and MIR syntax for expressing synchronization scopes has changed to use *syncscope("<scope>")*, where <scope> can be "singlethread" (this replaces *singlethread* keyword), or a target-specific name. As before, if the scope is not specified, it defaults to CrossThread/System scope. Implementation details: - Mapping from synchronization scope name/string to synchronization scope id is stored in LLVM context; - CrossThread/System and SingleThread scopes are pre-defined to efficiently check for known scopes without comparing strings; - Synchronization scope names are stored in SYNC_SCOPE_NAMES_BLOCK in the bitcode. Differential Revision: https://reviews.llvm.org/D21723 llvm-svn: 307722
2017-07-12 06:23:00 +08:00
%atomicrmw.umax = atomicrmw umax i32* %word, i32 21 syncscope("singlethread") monotonic
; CHECK: %atomicrmw.umax = atomicrmw umax i32* %word, i32 21 syncscope("singlethread") monotonic
%atomicrmw.umin = atomicrmw volatile umin i32* %word, i32 22 syncscope("singlethread") monotonic
; CHECK: %atomicrmw.umin = atomicrmw volatile umin i32* %word, i32 22 syncscope("singlethread") monotonic
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
fence acquire
; CHECK: fence acquire
fence release
; CHECK: fence release
fence acq_rel
; CHECK: fence acq_rel
Enhance synchscope representation OpenCL 2.0 introduces the notion of memory scopes in atomic operations to global and local memory. These scopes restrict how synchronization is achieved, which can result in improved performance. This change extends existing notion of synchronization scopes in LLVM to support arbitrary scopes expressed as target-specific strings, in addition to the already defined scopes (single thread, system). The LLVM IR and MIR syntax for expressing synchronization scopes has changed to use *syncscope("<scope>")*, where <scope> can be "singlethread" (this replaces *singlethread* keyword), or a target-specific name. As before, if the scope is not specified, it defaults to CrossThread/System scope. Implementation details: - Mapping from synchronization scope name/string to synchronization scope id is stored in LLVM context; - CrossThread/System and SingleThread scopes are pre-defined to efficiently check for known scopes without comparing strings; - Synchronization scope names are stored in SYNC_SCOPE_NAMES_BLOCK in the bitcode. Differential Revision: https://reviews.llvm.org/D21723 llvm-svn: 307722
2017-07-12 06:23:00 +08:00
fence syncscope("singlethread") seq_cst
; CHECK: fence syncscope("singlethread") seq_cst
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
%ld.1 = load atomic i32, i32* %word monotonic, align 4
; CHECK: %ld.1 = load atomic i32, i32* %word monotonic, align 4
%ld.2 = load atomic volatile i32, i32* %word acquire, align 8
; CHECK: %ld.2 = load atomic volatile i32, i32* %word acquire, align 8
Enhance synchscope representation OpenCL 2.0 introduces the notion of memory scopes in atomic operations to global and local memory. These scopes restrict how synchronization is achieved, which can result in improved performance. This change extends existing notion of synchronization scopes in LLVM to support arbitrary scopes expressed as target-specific strings, in addition to the already defined scopes (single thread, system). The LLVM IR and MIR syntax for expressing synchronization scopes has changed to use *syncscope("<scope>")*, where <scope> can be "singlethread" (this replaces *singlethread* keyword), or a target-specific name. As before, if the scope is not specified, it defaults to CrossThread/System scope. Implementation details: - Mapping from synchronization scope name/string to synchronization scope id is stored in LLVM context; - CrossThread/System and SingleThread scopes are pre-defined to efficiently check for known scopes without comparing strings; - Synchronization scope names are stored in SYNC_SCOPE_NAMES_BLOCK in the bitcode. Differential Revision: https://reviews.llvm.org/D21723 llvm-svn: 307722
2017-07-12 06:23:00 +08:00
%ld.3 = load atomic volatile i32, i32* %word syncscope("singlethread") seq_cst, align 16
; CHECK: %ld.3 = load atomic volatile i32, i32* %word syncscope("singlethread") seq_cst, align 16
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
store atomic i32 23, i32* %word monotonic, align 4
; CHECK: store atomic i32 23, i32* %word monotonic, align 4
store atomic volatile i32 24, i32* %word monotonic, align 4
; CHECK: store atomic volatile i32 24, i32* %word monotonic, align 4
Enhance synchscope representation OpenCL 2.0 introduces the notion of memory scopes in atomic operations to global and local memory. These scopes restrict how synchronization is achieved, which can result in improved performance. This change extends existing notion of synchronization scopes in LLVM to support arbitrary scopes expressed as target-specific strings, in addition to the already defined scopes (single thread, system). The LLVM IR and MIR syntax for expressing synchronization scopes has changed to use *syncscope("<scope>")*, where <scope> can be "singlethread" (this replaces *singlethread* keyword), or a target-specific name. As before, if the scope is not specified, it defaults to CrossThread/System scope. Implementation details: - Mapping from synchronization scope name/string to synchronization scope id is stored in LLVM context; - CrossThread/System and SingleThread scopes are pre-defined to efficiently check for known scopes without comparing strings; - Synchronization scope names are stored in SYNC_SCOPE_NAMES_BLOCK in the bitcode. Differential Revision: https://reviews.llvm.org/D21723 llvm-svn: 307722
2017-07-12 06:23:00 +08:00
store atomic volatile i32 25, i32* %word syncscope("singlethread") monotonic, align 4
; CHECK: store atomic volatile i32 25, i32* %word syncscope("singlethread") monotonic, align 4
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
ret void
}
Allow FP types for atomicrmw xchg llvm-svn: 351427
2019-01-17 18:49:01 +08:00
define void @fp_atomics(float* %word) {
; CHECK: %atomicrmw.xchg = atomicrmw xchg float* %word, float 1.000000e+00 monotonic
%atomicrmw.xchg = atomicrmw xchg float* %word, float 1.0 monotonic
Reapply "IR: Add fp operations to atomicrmw" This reapplies commits r351778 and r351782 with RISCV test fixes. llvm-svn: 351850
2019-01-23 02:18:02 +08:00
; CHECK: %atomicrmw.fadd = atomicrmw fadd float* %word, float 1.000000e+00 monotonic
%atomicrmw.fadd = atomicrmw fadd float* %word, float 1.0 monotonic
; CHECK: %atomicrmw.fsub = atomicrmw fsub float* %word, float 1.000000e+00 monotonic
%atomicrmw.fsub = atomicrmw fsub float* %word, float 1.0 monotonic
Allow FP types for atomicrmw xchg llvm-svn: 351427
2019-01-17 18:49:01 +08:00
ret void
}
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
;; Fast Math Flags
[IR] Add a dedicated FNeg IR Instruction The IEEE-754 Standard makes it clear that fneg(x) and fsub(-0.0, x) are two different operations. The former is a bitwise operation, while the latter is an arithmetic operation. This patch creates a dedicated FNeg IR Instruction to model that behavior. Differential Revision: https://reviews.llvm.org/D53877 llvm-svn: 346774
2018-11-14 02:15:47 +08:00
define void @fastmathflags_unop(float %op1) {
%f.nnan = fneg nnan float %op1
; CHECK: %f.nnan = fneg nnan float %op1
%f.ninf = fneg ninf float %op1
; CHECK: %f.ninf = fneg ninf float %op1
%f.nsz = fneg nsz float %op1
; CHECK: %f.nsz = fneg nsz float %op1
%f.arcp = fneg arcp float %op1
; CHECK: %f.arcp = fneg arcp float %op1
%f.contract = fneg contract float %op1
; CHECK: %f.contract = fneg contract float %op1
%f.afn = fneg afn float %op1
; CHECK: %f.afn = fneg afn float %op1
%f.reassoc = fneg reassoc float %op1
; CHECK: %f.reassoc = fneg reassoc float %op1
%f.fast = fneg fast float %op1
; CHECK: %f.fast = fneg fast float %op1
ret void
}
define void @fastmathflags_binops(float %op1, float %op2) {
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
%f.nnan = fadd nnan float %op1, %op2
; CHECK: %f.nnan = fadd nnan float %op1, %op2
%f.ninf = fadd ninf float %op1, %op2
; CHECK: %f.ninf = fadd ninf float %op1, %op2
%f.nsz = fadd nsz float %op1, %op2
; CHECK: %f.nsz = fadd nsz float %op1, %op2
%f.arcp = fadd arcp float %op1, %op2
; CHECK: %f.arcp = fadd arcp float %op1, %op2
[IR] Add AllowContract to FastMathFlags -ffp-contract=fast does not currently work with LTO because it's passed as a TargetOption to the backend rather than in the IR. This adds it to FastMathFlags. This is toward fixing PR25721 Differential Revision: https://reviews.llvm.org/D31164 llvm-svn: 298939
2017-03-29 04:11:52 +08:00
%f.contract = fadd contract float %op1, %op2
; CHECK: %f.contract = fadd contract float %op1, %op2
[IR] redefine 'UnsafeAlgebra' / 'reassoc' fast-math-flags and add 'trans' fast-math-flag As discussed on llvm-dev: http://lists.llvm.org/pipermail/llvm-dev/2016-November/107104.html and again more recently: http://lists.llvm.org/pipermail/llvm-dev/2017-October/118118.html ...this is a step in cleaning up our fast-math-flags implementation in IR to better match the capabilities of both clang's user-visible flags and the backend's flags for SDNode. As proposed in the above threads, we're replacing the 'UnsafeAlgebra' bit (which had the 'umbrella' meaning that all flags are set) with a new bit that only applies to algebraic reassociation - 'AllowReassoc'. We're also adding a bit to allow approximations for library functions called 'ApproxFunc' (this was initially proposed as 'libm' or similar). ...and we're out of bits. 7 bits ought to be enough for anyone, right? :) FWIW, I did look at getting this out of SubclassOptionalData via SubclassData (spacious 16-bits), but that's apparently already used for other purposes. Also, I don't think we can just add a field to FPMathOperator because Operator is not intended to be instantiated. We'll defer movement of FMF to another day. We keep the 'fast' keyword. I thought about removing that, but seeing IR like this: %f.fast = fadd reassoc nnan ninf nsz arcp contract afn float %op1, %op2 ...made me think we want to keep the shortcut synonym. Finally, this change is binary incompatible with existing IR as seen in the compatibility tests. This statement: "Newer releases can ignore features from older releases, but they cannot miscompile them. For example, if nsw is ever replaced with something else, dropping it would be a valid way to upgrade the IR." ( http://llvm.org/docs/DeveloperPolicy.html#ir-backwards-compatibility ) ...provides the flexibility we want to make this change without requiring a new IR version. Ie, we're not loosening the FP strictness of existing IR. At worst, we will fail to optimize some previously 'fast' code because it's no longer recognized as 'fast'. This should get fixed as we audit/squash all of the uses of 'isFast()'. Note: an inter-dependent clang commit to use the new API name should closely follow commit. Differential Revision: https://reviews.llvm.org/D39304 llvm-svn: 317488
2017-11-07 00:27:15 +08:00
%f.afn = fadd afn float %op1, %op2
; CHECK: %f.afn = fadd afn float %op1, %op2
%f.reassoc = fadd reassoc float %op1, %op2
; CHECK: %f.reassoc = fadd reassoc float %op1, %op2
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
%f.fast = fadd fast float %op1, %op2
; CHECK: %f.fast = fadd fast float %op1, %op2
ret void
}
[IR] allow fast-math-flags on select of FP values This is a minimal start to correcting a problem most directly discussed in PR38086: https://bugs.llvm.org/show_bug.cgi?id=38086 We have been hacking around a limitation for FP select patterns by using the fast-math-flags on the condition of the select rather than the select itself. This patch just allows FMF to appear with the 'select' opcode. No changes are needed to "FPMathOperator" because it already includes select-of-FP because that definition is based on the (return) value type. Once we have this ability, we can start correcting and adding IR transforms to use the FMF on a 'select' instruction. The instcombine and vectorizer test diffs only show that the IRBuilder change is behaving as expected by applying an FMF guard value to 'select'. For reference: rL241901 - allowed FMF with fcmp rL255555 - allowed FMF with FP calls Differential Revision: https://reviews.llvm.org/D61917 llvm-svn: 361401
2019-05-22 23:50:46 +08:00
define void @fastmathflags_select(i1 %cond, float %op1, float %op2) {
%f.nnan = select nnan i1 %cond, float %op1, float %op2
; CHECK: %f.nnan = select nnan i1 %cond, float %op1, float %op2
%f.ninf = select ninf i1 %cond, float %op1, float %op2
; CHECK: %f.ninf = select ninf i1 %cond, float %op1, float %op2
%f.nsz = select nsz i1 %cond, float %op1, float %op2
; CHECK: %f.nsz = select nsz i1 %cond, float %op1, float %op2
%f.arcp = select arcp i1 %cond, float %op1, float %op2
; CHECK: %f.arcp = select arcp i1 %cond, float %op1, float %op2
%f.contract = select contract i1 %cond, float %op1, float %op2
; CHECK: %f.contract = select contract i1 %cond, float %op1, float %op2
%f.afn = select afn i1 %cond, float %op1, float %op2
; CHECK: %f.afn = select afn i1 %cond, float %op1, float %op2
%f.reassoc = select reassoc i1 %cond, float %op1, float %op2
; CHECK: %f.reassoc = select reassoc i1 %cond, float %op1, float %op2
%f.fast = select fast i1 %cond, float %op1, float %op2
; CHECK: %f.fast = select fast i1 %cond, float %op1, float %op2
ret void
}
define void @fastmathflags_vector_select(<2 x i1> %cond, <2 x double> %op1, <2 x double> %op2) {
%f.nnan.nsz = select nnan nsz <2 x i1> %cond, <2 x double> %op1, <2 x double> %op2
; CHECK: %f.nnan.nsz = select nnan nsz <2 x i1> %cond, <2 x double> %op1, <2 x double> %op2
%f.fast = select fast <2 x i1> %cond, <2 x double> %op1, <2 x double> %op2
; CHECK: %f.fast = select fast <2 x i1> %cond, <2 x double> %op1, <2 x double> %op2
ret void
}
[IR] Allow fast math flags on calls with floating point array type. Summary: This extends the rules for when a call instruction is deemed to be an FPMathOperator, which is based on the type of the call (i.e. the return type of the function being called). Previously we only allowed floating-point and vector-of-floating-point types. Now we also allow arrays (nested to any depth) of floating-point and vector-of-floating-point types. This was motivated by llpc, the pipeline compiler for AMD GPUs (https://github.com/GPUOpen-Drivers/llpc). llpc has many math library functions that operate on vectors, typically represented as <4 x float>, and some that operate on matrices, typically represented as [4 x <4 x float>], and it's useful to be able to decorate calls to all of them with fast math flags. Reviewers: spatel, wristow, arsenm, hfinkel, aemerson, efriedma, cameron.mcinally, mcberg2017, jmolloy Subscribers: wdng, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D69161
2019-10-17 16:25:23 +08:00
define void @fastmathflags_array_select(i1 %cond, [2 x double] %op1, [2 x double] %op2) {
%f.nnan.nsz = select nnan nsz i1 %cond, [2 x double] %op1, [2 x double] %op2
; CHECK: %f.nnan.nsz = select nnan nsz i1 %cond, [2 x double] %op1, [2 x double] %op2
%f.fast = select fast i1 %cond, [2 x double] %op1, [2 x double] %op2
; CHECK: %f.fast = select fast i1 %cond, [2 x double] %op1, [2 x double] %op2
ret void
}
[IR] allow fast-math-flags on phi of FP values (2nd try) The changes here are based on the corresponding diffs for allowing FMF on 'select': D61917 <https://reviews.llvm.org/D61917> As discussed there, we want to have fast-math-flags be a property of an FP value because the alternative (having them on things like fcmp) leads to logical inconsistency such as: https://bugs.llvm.org/show_bug.cgi?id=38086 The earlier patch for select made almost no practical difference because most unoptimized conditional code begins life as a phi (based on what I see in clang). Similarly, I don't expect this patch to do much on its own either because SimplifyCFG promptly drops the flags when converting to select on a minimal example like: https://bugs.llvm.org/show_bug.cgi?id=39535 But once we have this plumbing in place, we should be able to wire up the FMF propagation and start solving cases like that. The change to RecurrenceDescriptor::AddReductionVar() is required to prevent a regression in a LoopVectorize test. We are intersecting the FMF of any FPMathOperator there, so if a phi is not properly annotated, new math instructions may not be either. Once we fix the propagation in SimplifyCFG, it may be safe to remove that hack. Differential Revision: https://reviews.llvm.org/D67564 llvm-svn: 372878
2019-09-25 22:35:02 +08:00
define void @fastmathflags_phi(i1 %cond, float %f1, float %f2, double %d1, double %d2, half %h1, half %h2) {
entry:
br i1 %cond, label %L1, label %L2
L1:
br label %exit
L2:
br label %exit
exit:
%p.nnan = phi nnan float [ %f1, %L1 ], [ %f2, %L2 ]
; CHECK: %p.nnan = phi nnan float [ %f1, %L1 ], [ %f2, %L2 ]
%p.ninf = phi ninf double [ %d1, %L1 ], [ %d2, %L2 ]
; CHECK: %p.ninf = phi ninf double [ %d1, %L1 ], [ %d2, %L2 ]
%p.contract = phi contract half [ %h1, %L1 ], [ %h2, %L2 ]
; CHECK: %p.contract = phi contract half [ %h1, %L1 ], [ %h2, %L2 ]
%p.nsz.reassoc = phi reassoc nsz float [ %f1, %L1 ], [ %f2, %L2 ]
; CHECK: %p.nsz.reassoc = phi reassoc nsz float [ %f1, %L1 ], [ %f2, %L2 ]
%p.fast = phi fast half [ %h2, %L1 ], [ %h1, %L2 ]
; CHECK: %p.fast = phi fast half [ %h2, %L1 ], [ %h1, %L2 ]
ret void
}
Add FMF to vector ops for phi Summary: Small amendment to handle vector cases for D67564. Reviewers: spatel, eli.friedman, hfinkel, cameron.mcinally, arsenm, jmolloy, bogner Reviewed By: cameron.mcinally, bogner Subscribers: llvm-commits, efriedma, reames, bogner, wdng Differential Revision: https://reviews.llvm.org/D68748 llvm-svn: 374794
2019-10-15 01:39:32 +08:00
define void @fastmathflags_vector_phi(i1 %cond, <4 x float> %f1, <4 x float> %f2, <2 x double> %d1, <2 x double> %d2, <8 x half> %h1, <8 x half> %h2) {
entry:
br i1 %cond, label %L1, label %L2
L1:
br label %exit
L2:
br label %exit
exit:
%p.nnan = phi nnan <4 x float> [ %f1, %L1 ], [ %f2, %L2 ]
; CHECK: %p.nnan = phi nnan <4 x float> [ %f1, %L1 ], [ %f2, %L2 ]
%p.ninf = phi ninf <2 x double> [ %d1, %L1 ], [ %d2, %L2 ]
; CHECK: %p.ninf = phi ninf <2 x double> [ %d1, %L1 ], [ %d2, %L2 ]
%p.contract = phi contract <8 x half> [ %h1, %L1 ], [ %h2, %L2 ]
; CHECK: %p.contract = phi contract <8 x half> [ %h1, %L1 ], [ %h2, %L2 ]
%p.nsz.reassoc = phi reassoc nsz <4 x float> [ %f1, %L1 ], [ %f2, %L2 ]
; CHECK: %p.nsz.reassoc = phi reassoc nsz <4 x float> [ %f1, %L1 ], [ %f2, %L2 ]
%p.fast = phi fast <8 x half> [ %h2, %L1 ], [ %h1, %L2 ]
; CHECK: %p.fast = phi fast <8 x half> [ %h2, %L1 ], [ %h1, %L2 ]
ret void
}
[IR] Allow fast math flags on calls with floating point array type. Summary: This extends the rules for when a call instruction is deemed to be an FPMathOperator, which is based on the type of the call (i.e. the return type of the function being called). Previously we only allowed floating-point and vector-of-floating-point types. Now we also allow arrays (nested to any depth) of floating-point and vector-of-floating-point types. This was motivated by llpc, the pipeline compiler for AMD GPUs (https://github.com/GPUOpen-Drivers/llpc). llpc has many math library functions that operate on vectors, typically represented as <4 x float>, and some that operate on matrices, typically represented as [4 x <4 x float>], and it's useful to be able to decorate calls to all of them with fast math flags. Reviewers: spatel, wristow, arsenm, hfinkel, aemerson, efriedma, cameron.mcinally, mcberg2017, jmolloy Subscribers: wdng, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D69161
2019-10-17 16:25:23 +08:00
define void @fastmathflags_array_phi(i1 %cond, [4 x float] %f1, [4 x float] %f2, [2 x double] %d1, [2 x double] %d2, [8 x half] %h1, [8 x half] %h2) {
entry:
br i1 %cond, label %L1, label %L2
L1:
br label %exit
L2:
br label %exit
exit:
%p.nnan = phi nnan [4 x float] [ %f1, %L1 ], [ %f2, %L2 ]
; CHECK: %p.nnan = phi nnan [4 x float] [ %f1, %L1 ], [ %f2, %L2 ]
%p.ninf = phi ninf [2 x double] [ %d1, %L1 ], [ %d2, %L2 ]
; CHECK: %p.ninf = phi ninf [2 x double] [ %d1, %L1 ], [ %d2, %L2 ]
%p.contract = phi contract [8 x half] [ %h1, %L1 ], [ %h2, %L2 ]
; CHECK: %p.contract = phi contract [8 x half] [ %h1, %L1 ], [ %h2, %L2 ]
%p.nsz.reassoc = phi reassoc nsz [4 x float] [ %f1, %L1 ], [ %f2, %L2 ]
; CHECK: %p.nsz.reassoc = phi reassoc nsz [4 x float] [ %f1, %L1 ], [ %f2, %L2 ]
%p.fast = phi fast [8 x half] [ %h2, %L1 ], [ %h1, %L2 ]
; CHECK: %p.fast = phi fast [8 x half] [ %h2, %L1 ], [ %h1, %L2 ]
ret void
}
add fast-math-flags to 'call' instructions (PR21290) This patch adds optional fast-math-flags (the same that apply to fmul/fadd/fsub/fdiv/frem/fcmp) to call instructions in IR. Follow-up patches would use these flags in LibCallSimplifier, add support to clang, and extend FMF to the DAG for calls. Motivating example: %y = fmul fast float %x, %x %z = tail call float @sqrtf(float %y) We'd like to be able to optimize sqrt(x*x) into fabs(x). We do this today using a function-wide attribute for unsafe-math, but we really want to trigger on the instructions themselves: %z = tail call fast float @sqrtf(float %y) because in an LTO build it's possible that calls with fast semantics have been inlined into a function with non-fast semantics. The code changes and tests are based on the recent commits that added "notail": http://reviews.llvm.org/rL252368 and added FMF to fcmp: http://reviews.llvm.org/rL241901 Differential Revision: http://reviews.llvm.org/D14707 llvm-svn: 255555
2015-12-15 05:59:03 +08:00
; Check various fast math flags and floating-point types on calls.
[IR] Allow fast math flags on calls with floating point array type. Summary: This extends the rules for when a call instruction is deemed to be an FPMathOperator, which is based on the type of the call (i.e. the return type of the function being called). Previously we only allowed floating-point and vector-of-floating-point types. Now we also allow arrays (nested to any depth) of floating-point and vector-of-floating-point types. This was motivated by llpc, the pipeline compiler for AMD GPUs (https://github.com/GPUOpen-Drivers/llpc). llpc has many math library functions that operate on vectors, typically represented as <4 x float>, and some that operate on matrices, typically represented as [4 x <4 x float>], and it's useful to be able to decorate calls to all of them with fast math flags. Reviewers: spatel, wristow, arsenm, hfinkel, aemerson, efriedma, cameron.mcinally, mcberg2017, jmolloy Subscribers: wdng, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D69161
2019-10-17 16:25:23 +08:00
declare float @fmf_f32()
declare double @fmf_f64()
declare <4 x double> @fmf_v4f64()
add fast-math-flags to 'call' instructions (PR21290) This patch adds optional fast-math-flags (the same that apply to fmul/fadd/fsub/fdiv/frem/fcmp) to call instructions in IR. Follow-up patches would use these flags in LibCallSimplifier, add support to clang, and extend FMF to the DAG for calls. Motivating example: %y = fmul fast float %x, %x %z = tail call float @sqrtf(float %y) We'd like to be able to optimize sqrt(x*x) into fabs(x). We do this today using a function-wide attribute for unsafe-math, but we really want to trigger on the instructions themselves: %z = tail call fast float @sqrtf(float %y) because in an LTO build it's possible that calls with fast semantics have been inlined into a function with non-fast semantics. The code changes and tests are based on the recent commits that added "notail": http://reviews.llvm.org/rL252368 and added FMF to fcmp: http://reviews.llvm.org/rL241901 Differential Revision: http://reviews.llvm.org/D14707 llvm-svn: 255555
2015-12-15 05:59:03 +08:00
; CHECK-LABEL: fastMathFlagsForCalls(
define void @fastMathFlagsForCalls(float %f, double %d1, <4 x double> %d2) {
[IR] Allow fast math flags on calls with floating point array type. Summary: This extends the rules for when a call instruction is deemed to be an FPMathOperator, which is based on the type of the call (i.e. the return type of the function being called). Previously we only allowed floating-point and vector-of-floating-point types. Now we also allow arrays (nested to any depth) of floating-point and vector-of-floating-point types. This was motivated by llpc, the pipeline compiler for AMD GPUs (https://github.com/GPUOpen-Drivers/llpc). llpc has many math library functions that operate on vectors, typically represented as <4 x float>, and some that operate on matrices, typically represented as [4 x <4 x float>], and it's useful to be able to decorate calls to all of them with fast math flags. Reviewers: spatel, wristow, arsenm, hfinkel, aemerson, efriedma, cameron.mcinally, mcberg2017, jmolloy Subscribers: wdng, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D69161
2019-10-17 16:25:23 +08:00
%call.fast = call fast float @fmf_f32()
; CHECK: %call.fast = call fast float @fmf_f32()
; Throw in some other attributes to make sure those stay in the right places.
%call.nsz.arcp = notail call nsz arcp double @fmf_f64()
; CHECK: %call.nsz.arcp = notail call nsz arcp double @fmf_f64()
%call.nnan.ninf = tail call nnan ninf fastcc <4 x double> @fmf_v4f64()
; CHECK: %call.nnan.ninf = tail call nnan ninf fastcc <4 x double> @fmf_v4f64()
ret void
}
declare [2 x float] @fmf_a2f32()
declare [2 x double] @fmf_a2f64()
declare [2 x <4 x double>] @fmf_a2v4f64()
; CHECK-LABEL: fastMathFlagsForArrayCalls(
define void @fastMathFlagsForArrayCalls([2 x float] %f, [2 x double] %d1, [2 x <4 x double>] %d2) {
%call.fast = call fast [2 x float] @fmf_a2f32()
; CHECK: %call.fast = call fast [2 x float] @fmf_a2f32()
add fast-math-flags to 'call' instructions (PR21290) This patch adds optional fast-math-flags (the same that apply to fmul/fadd/fsub/fdiv/frem/fcmp) to call instructions in IR. Follow-up patches would use these flags in LibCallSimplifier, add support to clang, and extend FMF to the DAG for calls. Motivating example: %y = fmul fast float %x, %x %z = tail call float @sqrtf(float %y) We'd like to be able to optimize sqrt(x*x) into fabs(x). We do this today using a function-wide attribute for unsafe-math, but we really want to trigger on the instructions themselves: %z = tail call fast float @sqrtf(float %y) because in an LTO build it's possible that calls with fast semantics have been inlined into a function with non-fast semantics. The code changes and tests are based on the recent commits that added "notail": http://reviews.llvm.org/rL252368 and added FMF to fcmp: http://reviews.llvm.org/rL241901 Differential Revision: http://reviews.llvm.org/D14707 llvm-svn: 255555
2015-12-15 05:59:03 +08:00
; Throw in some other attributes to make sure those stay in the right places.
[IR] Allow fast math flags on calls with floating point array type. Summary: This extends the rules for when a call instruction is deemed to be an FPMathOperator, which is based on the type of the call (i.e. the return type of the function being called). Previously we only allowed floating-point and vector-of-floating-point types. Now we also allow arrays (nested to any depth) of floating-point and vector-of-floating-point types. This was motivated by llpc, the pipeline compiler for AMD GPUs (https://github.com/GPUOpen-Drivers/llpc). llpc has many math library functions that operate on vectors, typically represented as <4 x float>, and some that operate on matrices, typically represented as [4 x <4 x float>], and it's useful to be able to decorate calls to all of them with fast math flags. Reviewers: spatel, wristow, arsenm, hfinkel, aemerson, efriedma, cameron.mcinally, mcberg2017, jmolloy Subscribers: wdng, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D69161
2019-10-17 16:25:23 +08:00
%call.nsz.arcp = notail call nsz arcp [2 x double] @fmf_a2f64()
; CHECK: %call.nsz.arcp = notail call nsz arcp [2 x double] @fmf_a2f64()
add fast-math-flags to 'call' instructions (PR21290) This patch adds optional fast-math-flags (the same that apply to fmul/fadd/fsub/fdiv/frem/fcmp) to call instructions in IR. Follow-up patches would use these flags in LibCallSimplifier, add support to clang, and extend FMF to the DAG for calls. Motivating example: %y = fmul fast float %x, %x %z = tail call float @sqrtf(float %y) We'd like to be able to optimize sqrt(x*x) into fabs(x). We do this today using a function-wide attribute for unsafe-math, but we really want to trigger on the instructions themselves: %z = tail call fast float @sqrtf(float %y) because in an LTO build it's possible that calls with fast semantics have been inlined into a function with non-fast semantics. The code changes and tests are based on the recent commits that added "notail": http://reviews.llvm.org/rL252368 and added FMF to fcmp: http://reviews.llvm.org/rL241901 Differential Revision: http://reviews.llvm.org/D14707 llvm-svn: 255555
2015-12-15 05:59:03 +08:00
[IR] Allow fast math flags on calls with floating point array type. Summary: This extends the rules for when a call instruction is deemed to be an FPMathOperator, which is based on the type of the call (i.e. the return type of the function being called). Previously we only allowed floating-point and vector-of-floating-point types. Now we also allow arrays (nested to any depth) of floating-point and vector-of-floating-point types. This was motivated by llpc, the pipeline compiler for AMD GPUs (https://github.com/GPUOpen-Drivers/llpc). llpc has many math library functions that operate on vectors, typically represented as <4 x float>, and some that operate on matrices, typically represented as [4 x <4 x float>], and it's useful to be able to decorate calls to all of them with fast math flags. Reviewers: spatel, wristow, arsenm, hfinkel, aemerson, efriedma, cameron.mcinally, mcberg2017, jmolloy Subscribers: wdng, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D69161
2019-10-17 16:25:23 +08:00
%call.nnan.ninf = tail call nnan ninf fastcc [2 x <4 x double>] @fmf_a2v4f64()
; CHECK: %call.nnan.ninf = tail call nnan ninf fastcc [2 x <4 x double>] @fmf_a2v4f64()
add fast-math-flags to 'call' instructions (PR21290) This patch adds optional fast-math-flags (the same that apply to fmul/fadd/fsub/fdiv/frem/fcmp) to call instructions in IR. Follow-up patches would use these flags in LibCallSimplifier, add support to clang, and extend FMF to the DAG for calls. Motivating example: %y = fmul fast float %x, %x %z = tail call float @sqrtf(float %y) We'd like to be able to optimize sqrt(x*x) into fabs(x). We do this today using a function-wide attribute for unsafe-math, but we really want to trigger on the instructions themselves: %z = tail call fast float @sqrtf(float %y) because in an LTO build it's possible that calls with fast semantics have been inlined into a function with non-fast semantics. The code changes and tests are based on the recent commits that added "notail": http://reviews.llvm.org/rL252368 and added FMF to fcmp: http://reviews.llvm.org/rL241901 Differential Revision: http://reviews.llvm.org/D14707 llvm-svn: 255555
2015-12-15 05:59:03 +08:00
ret void
}
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
;; Type System
%opaquety = type opaque
define void @typesystem() {
%p0 = bitcast i8* null to i32 (i32)*
; CHECK: %p0 = bitcast i8* null to i32 (i32)*
%p1 = bitcast i8* null to void (i8*)*
; CHECK: %p1 = bitcast i8* null to void (i8*)*
%p2 = bitcast i8* null to i32 (i8*, ...)*
; CHECK: %p2 = bitcast i8* null to i32 (i8*, ...)*
%p3 = bitcast i8* null to { i32, i8 } (i8*, ...)*
; CHECK: %p3 = bitcast i8* null to { i32, i8 } (i8*, ...)*
%p4 = bitcast i8* null to <{ i32, i8 }> (i8*, ...)*
; CHECK: %p4 = bitcast i8* null to <{ i32, i8 }> (i8*, ...)*
%p5 = bitcast i8* null to <{ i32, i8 }> (<{ i8*, i64 }>*, ...)*
; CHECK: %p5 = bitcast i8* null to <{ i32, i8 }> (<{ i8*, i64 }>*, ...)*
%t0 = alloca i1942652
; CHECK: %t0 = alloca i1942652
%t1 = alloca half
; CHECK: %t1 = alloca half
%t2 = alloca float
; CHECK: %t2 = alloca float
%t3 = alloca double
; CHECK: %t3 = alloca double
%t4 = alloca fp128
; CHECK: %t4 = alloca fp128
%t5 = alloca x86_fp80
; CHECK: %t5 = alloca x86_fp80
%t6 = alloca ppc_fp128
; CHECK: %t6 = alloca ppc_fp128
%t7 = alloca x86_mmx
; CHECK: %t7 = alloca x86_mmx
%t8 = alloca %opaquety*
; CHECK: %t8 = alloca %opaquety*
Scalable Vector IR Type with further LTO fixes Reintroduces the scalable vector IR type from D32530, after it was reverted a couple of times due to increasing chromium LTO build times. This latest incarnation removes the walk over aggregate types from the verifier entirely, in favor of rejecting scalable vectors in the isValidElementType methods in ArrayType and StructType. This removes the 70% degradation observed with the second repro tarball from PR42210. Reviewers: thakis, hans, rengolin, sdesmalen Reviewed By: sdesmalen Differential Revision: https://reviews.llvm.org/D64079 llvm-svn: 365203
2019-07-05 20:48:16 +08:00
%t9 = alloca <4 x i32>
; CHECK: %t9 = alloca <4 x i32>
%t10 = alloca <vscale x 4 x i32>
; CHECK: %t10 = alloca <vscale x 4 x i32>
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
ret void
}
[Bitcode] Add compatibility tests for new instructions Adds basic compatibility tests for the following instructions: catchpad, catchendpad, cleanuppad, cleanupendpad, terminatepad, cleanupret, catchret llvm-svn: 247087
2015-09-09 06:33:23 +08:00
declare void @llvm.token(token)
; CHECK: declare void @llvm.token(token)
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
;; Inline Assembler Expressions
define void @inlineasm(i32 %arg) {
call i32 asm "bswap $0", "=r,r"(i32 %arg)
; CHECK: call i32 asm "bswap $0", "=r,r"(i32 %arg)
call i32 asm sideeffect "blt $1, $2, $3", "=r,r,rm"(i32 %arg, i32 %arg)
; CHECK: call i32 asm sideeffect "blt $1, $2, $3", "=r,r,rm"(i32 %arg, i32 %arg)
ret void
}
;; Instructions
; Instructions -- Terminators
define void @instructions.terminators(i8 %val) personality i32 -10 {
br i1 false, label %iftrue, label %iffalse
; CHECK: br i1 false, label %iftrue, label %iffalse
br label %iftrue
; CHECK: br label %iftrue
iftrue:
ret void
; CHECK: ret void
iffalse:
switch i8 %val, label %defaultdest [
; CHECK: switch i8 %val, label %defaultdest [
i8 0, label %defaultdest.0
; CHECK: i8 0, label %defaultdest.0
i8 1, label %defaultdest.1
; CHECK: i8 1, label %defaultdest.1
i8 2, label %defaultdest.2
; CHECK: i8 2, label %defaultdest.2
]
; CHECK: ]
defaultdest:
ret void
defaultdest.0:
ret void
defaultdest.1:
ret void
defaultdest.2:
indirectbr i8* blockaddress(@instructions.terminators, %defaultdest.2), [label %defaultdest.2]
; CHECK: indirectbr i8* blockaddress(@instructions.terminators, %defaultdest.2), [label %defaultdest.2]
indirectbr i8* blockaddress(@instructions.terminators, %defaultdest.2), [label %defaultdest.2, label %defaultdest.2]
; CHECK: indirectbr i8* blockaddress(@instructions.terminators, %defaultdest.2), [label %defaultdest.2, label %defaultdest.2]
invoke fastcc void @f.fastcc()
; CHECK: invoke fastcc void @f.fastcc()
to label %defaultdest unwind label %exc
; CHECK: to label %defaultdest unwind label %exc
exc:
%cleanup = landingpad i32 cleanup
resume i32 undef
; CHECK: resume i32 undef
unreachable
; CHECK: unreachable
ret void
}
[Bitcode] Add compatibility tests for new instructions Adds basic compatibility tests for the following instructions: catchpad, catchendpad, cleanuppad, cleanupendpad, terminatepad, cleanupret, catchret llvm-svn: 247087
2015-09-09 06:33:23 +08:00
define i32 @instructions.win_eh.1() personality i32 -3 {
entry:
%arg1 = alloca i32
%arg2 = alloca i32
[IR] Reformulate LLVM's EH funclet IR While we have successfully implemented a funclet-oriented EH scheme on top of LLVM IR, our scheme has some notable deficiencies: - catchendpad and cleanupendpad are necessary in the current design but they are difficult to explain to others, even to seasoned LLVM experts. - catchendpad and cleanupendpad are optimization barriers. They cannot be split and force all potentially throwing call-sites to be invokes. This has a noticable effect on the quality of our code generation. - catchpad, while similar in some aspects to invoke, is fairly awkward. It is unsplittable, starts a funclet, and has control flow to other funclets. - The nesting relationship between funclets is currently a property of control flow edges. Because of this, we are forced to carefully analyze the flow graph to see if there might potentially exist illegal nesting among funclets. While we have logic to clone funclets when they are illegally nested, it would be nicer if we had a representation which forbade them upfront. Let's clean this up a bit by doing the following: - Instead, make catchpad more like cleanuppad and landingpad: no control flow, just a bunch of simple operands; catchpad would be splittable. - Introduce catchswitch, a control flow instruction designed to model the constraints of funclet oriented EH. - Make funclet scoping explicit by having funclet instructions consume the token produced by the funclet which contains them. - Remove catchendpad and cleanupendpad. Their presence can be inferred implicitly using coloring information. N.B. The state numbering code for the CLR has been updated but the veracity of it's output cannot be spoken for. An expert should take a look to make sure the results are reasonable. Reviewers: rnk, JosephTremoulet, andrew.w.kaylor Differential Revision: http://reviews.llvm.org/D15139 llvm-svn: 255422
2015-12-12 13:38:55 +08:00
invoke void @f.ccc() to label %normal unwind label %catchswitch1
invoke void @f.ccc() to label %normal unwind label %catchswitch2
invoke void @f.ccc() to label %normal unwind label %catchswitch3
catchswitch1:
[IR] Remove terminatepad It turns out that terminatepad gives little benefit over a cleanuppad which calls the termination function. This is not sufficient to implement fully generic filters but MSVC doesn't support them which makes terminatepad a little over-designed. Depends on D15478. Differential Revision: http://reviews.llvm.org/D15479 llvm-svn: 255522
2015-12-15 02:34:23 +08:00
%cs1 = catchswitch within none [label %catchpad1] unwind to caller
[Bitcode] Add compatibility tests for new instructions Adds basic compatibility tests for the following instructions: catchpad, catchendpad, cleanuppad, cleanupendpad, terminatepad, cleanupret, catchret llvm-svn: 247087
2015-09-09 06:33:23 +08:00
catchpad1:
[IR] Reformulate LLVM's EH funclet IR While we have successfully implemented a funclet-oriented EH scheme on top of LLVM IR, our scheme has some notable deficiencies: - catchendpad and cleanupendpad are necessary in the current design but they are difficult to explain to others, even to seasoned LLVM experts. - catchendpad and cleanupendpad are optimization barriers. They cannot be split and force all potentially throwing call-sites to be invokes. This has a noticable effect on the quality of our code generation. - catchpad, while similar in some aspects to invoke, is fairly awkward. It is unsplittable, starts a funclet, and has control flow to other funclets. - The nesting relationship between funclets is currently a property of control flow edges. Because of this, we are forced to carefully analyze the flow graph to see if there might potentially exist illegal nesting among funclets. While we have logic to clone funclets when they are illegally nested, it would be nicer if we had a representation which forbade them upfront. Let's clean this up a bit by doing the following: - Instead, make catchpad more like cleanuppad and landingpad: no control flow, just a bunch of simple operands; catchpad would be splittable. - Introduce catchswitch, a control flow instruction designed to model the constraints of funclet oriented EH. - Make funclet scoping explicit by having funclet instructions consume the token produced by the funclet which contains them. - Remove catchendpad and cleanupendpad. Their presence can be inferred implicitly using coloring information. N.B. The state numbering code for the CLR has been updated but the veracity of it's output cannot be spoken for. An expert should take a look to make sure the results are reasonable. Reviewers: rnk, JosephTremoulet, andrew.w.kaylor Differential Revision: http://reviews.llvm.org/D15139 llvm-svn: 255422
2015-12-12 13:38:55 +08:00
catchpad within %cs1 []
br label %normal
; CHECK: catchpad within %cs1 []
; CHECK-NEXT: br label %normal
[Bitcode] Add compatibility tests for new instructions Adds basic compatibility tests for the following instructions: catchpad, catchendpad, cleanuppad, cleanupendpad, terminatepad, cleanupret, catchret llvm-svn: 247087
2015-09-09 06:33:23 +08:00
[IR] Reformulate LLVM's EH funclet IR While we have successfully implemented a funclet-oriented EH scheme on top of LLVM IR, our scheme has some notable deficiencies: - catchendpad and cleanupendpad are necessary in the current design but they are difficult to explain to others, even to seasoned LLVM experts. - catchendpad and cleanupendpad are optimization barriers. They cannot be split and force all potentially throwing call-sites to be invokes. This has a noticable effect on the quality of our code generation. - catchpad, while similar in some aspects to invoke, is fairly awkward. It is unsplittable, starts a funclet, and has control flow to other funclets. - The nesting relationship between funclets is currently a property of control flow edges. Because of this, we are forced to carefully analyze the flow graph to see if there might potentially exist illegal nesting among funclets. While we have logic to clone funclets when they are illegally nested, it would be nicer if we had a representation which forbade them upfront. Let's clean this up a bit by doing the following: - Instead, make catchpad more like cleanuppad and landingpad: no control flow, just a bunch of simple operands; catchpad would be splittable. - Introduce catchswitch, a control flow instruction designed to model the constraints of funclet oriented EH. - Make funclet scoping explicit by having funclet instructions consume the token produced by the funclet which contains them. - Remove catchendpad and cleanupendpad. Their presence can be inferred implicitly using coloring information. N.B. The state numbering code for the CLR has been updated but the veracity of it's output cannot be spoken for. An expert should take a look to make sure the results are reasonable. Reviewers: rnk, JosephTremoulet, andrew.w.kaylor Differential Revision: http://reviews.llvm.org/D15139 llvm-svn: 255422
2015-12-12 13:38:55 +08:00
catchswitch2:
%cs2 = catchswitch within none [label %catchpad2] unwind to caller
[Bitcode] Add compatibility tests for new instructions Adds basic compatibility tests for the following instructions: catchpad, catchendpad, cleanuppad, cleanupendpad, terminatepad, cleanupret, catchret llvm-svn: 247087
2015-09-09 06:33:23 +08:00
[IR] Reformulate LLVM's EH funclet IR While we have successfully implemented a funclet-oriented EH scheme on top of LLVM IR, our scheme has some notable deficiencies: - catchendpad and cleanupendpad are necessary in the current design but they are difficult to explain to others, even to seasoned LLVM experts. - catchendpad and cleanupendpad are optimization barriers. They cannot be split and force all potentially throwing call-sites to be invokes. This has a noticable effect on the quality of our code generation. - catchpad, while similar in some aspects to invoke, is fairly awkward. It is unsplittable, starts a funclet, and has control flow to other funclets. - The nesting relationship between funclets is currently a property of control flow edges. Because of this, we are forced to carefully analyze the flow graph to see if there might potentially exist illegal nesting among funclets. While we have logic to clone funclets when they are illegally nested, it would be nicer if we had a representation which forbade them upfront. Let's clean this up a bit by doing the following: - Instead, make catchpad more like cleanuppad and landingpad: no control flow, just a bunch of simple operands; catchpad would be splittable. - Introduce catchswitch, a control flow instruction designed to model the constraints of funclet oriented EH. - Make funclet scoping explicit by having funclet instructions consume the token produced by the funclet which contains them. - Remove catchendpad and cleanupendpad. Their presence can be inferred implicitly using coloring information. N.B. The state numbering code for the CLR has been updated but the veracity of it's output cannot be spoken for. An expert should take a look to make sure the results are reasonable. Reviewers: rnk, JosephTremoulet, andrew.w.kaylor Differential Revision: http://reviews.llvm.org/D15139 llvm-svn: 255422
2015-12-12 13:38:55 +08:00
catchpad2:
catchpad within %cs2 [i32* %arg1]
br label %normal
; CHECK: catchpad within %cs2 [i32* %arg1]
; CHECK-NEXT: br label %normal
[Bitcode] Add compatibility tests for new instructions Adds basic compatibility tests for the following instructions: catchpad, catchendpad, cleanuppad, cleanupendpad, terminatepad, cleanupret, catchret llvm-svn: 247087
2015-09-09 06:33:23 +08:00
[IR] Reformulate LLVM's EH funclet IR While we have successfully implemented a funclet-oriented EH scheme on top of LLVM IR, our scheme has some notable deficiencies: - catchendpad and cleanupendpad are necessary in the current design but they are difficult to explain to others, even to seasoned LLVM experts. - catchendpad and cleanupendpad are optimization barriers. They cannot be split and force all potentially throwing call-sites to be invokes. This has a noticable effect on the quality of our code generation. - catchpad, while similar in some aspects to invoke, is fairly awkward. It is unsplittable, starts a funclet, and has control flow to other funclets. - The nesting relationship between funclets is currently a property of control flow edges. Because of this, we are forced to carefully analyze the flow graph to see if there might potentially exist illegal nesting among funclets. While we have logic to clone funclets when they are illegally nested, it would be nicer if we had a representation which forbade them upfront. Let's clean this up a bit by doing the following: - Instead, make catchpad more like cleanuppad and landingpad: no control flow, just a bunch of simple operands; catchpad would be splittable. - Introduce catchswitch, a control flow instruction designed to model the constraints of funclet oriented EH. - Make funclet scoping explicit by having funclet instructions consume the token produced by the funclet which contains them. - Remove catchendpad and cleanupendpad. Their presence can be inferred implicitly using coloring information. N.B. The state numbering code for the CLR has been updated but the veracity of it's output cannot be spoken for. An expert should take a look to make sure the results are reasonable. Reviewers: rnk, JosephTremoulet, andrew.w.kaylor Differential Revision: http://reviews.llvm.org/D15139 llvm-svn: 255422
2015-12-12 13:38:55 +08:00
catchswitch3:
%cs3 = catchswitch within none [label %catchpad3] unwind label %cleanuppad1
[Bitcode] Add compatibility tests for new instructions Adds basic compatibility tests for the following instructions: catchpad, catchendpad, cleanuppad, cleanupendpad, terminatepad, cleanupret, catchret llvm-svn: 247087
2015-09-09 06:33:23 +08:00
[IR] Reformulate LLVM's EH funclet IR While we have successfully implemented a funclet-oriented EH scheme on top of LLVM IR, our scheme has some notable deficiencies: - catchendpad and cleanupendpad are necessary in the current design but they are difficult to explain to others, even to seasoned LLVM experts. - catchendpad and cleanupendpad are optimization barriers. They cannot be split and force all potentially throwing call-sites to be invokes. This has a noticable effect on the quality of our code generation. - catchpad, while similar in some aspects to invoke, is fairly awkward. It is unsplittable, starts a funclet, and has control flow to other funclets. - The nesting relationship between funclets is currently a property of control flow edges. Because of this, we are forced to carefully analyze the flow graph to see if there might potentially exist illegal nesting among funclets. While we have logic to clone funclets when they are illegally nested, it would be nicer if we had a representation which forbade them upfront. Let's clean this up a bit by doing the following: - Instead, make catchpad more like cleanuppad and landingpad: no control flow, just a bunch of simple operands; catchpad would be splittable. - Introduce catchswitch, a control flow instruction designed to model the constraints of funclet oriented EH. - Make funclet scoping explicit by having funclet instructions consume the token produced by the funclet which contains them. - Remove catchendpad and cleanupendpad. Their presence can be inferred implicitly using coloring information. N.B. The state numbering code for the CLR has been updated but the veracity of it's output cannot be spoken for. An expert should take a look to make sure the results are reasonable. Reviewers: rnk, JosephTremoulet, andrew.w.kaylor Differential Revision: http://reviews.llvm.org/D15139 llvm-svn: 255422
2015-12-12 13:38:55 +08:00
catchpad3:
catchpad within %cs3 [i32* %arg1, i32* %arg2]
br label %normal
; CHECK: catchpad within %cs3 [i32* %arg1, i32* %arg2]
; CHECK-NEXT: br label %normal
[Bitcode] Add compatibility tests for new instructions Adds basic compatibility tests for the following instructions: catchpad, catchendpad, cleanuppad, cleanupendpad, terminatepad, cleanupret, catchret llvm-svn: 247087
2015-09-09 06:33:23 +08:00
cleanuppad1:
[IR] Reformulate LLVM's EH funclet IR While we have successfully implemented a funclet-oriented EH scheme on top of LLVM IR, our scheme has some notable deficiencies: - catchendpad and cleanupendpad are necessary in the current design but they are difficult to explain to others, even to seasoned LLVM experts. - catchendpad and cleanupendpad are optimization barriers. They cannot be split and force all potentially throwing call-sites to be invokes. This has a noticable effect on the quality of our code generation. - catchpad, while similar in some aspects to invoke, is fairly awkward. It is unsplittable, starts a funclet, and has control flow to other funclets. - The nesting relationship between funclets is currently a property of control flow edges. Because of this, we are forced to carefully analyze the flow graph to see if there might potentially exist illegal nesting among funclets. While we have logic to clone funclets when they are illegally nested, it would be nicer if we had a representation which forbade them upfront. Let's clean this up a bit by doing the following: - Instead, make catchpad more like cleanuppad and landingpad: no control flow, just a bunch of simple operands; catchpad would be splittable. - Introduce catchswitch, a control flow instruction designed to model the constraints of funclet oriented EH. - Make funclet scoping explicit by having funclet instructions consume the token produced by the funclet which contains them. - Remove catchendpad and cleanupendpad. Their presence can be inferred implicitly using coloring information. N.B. The state numbering code for the CLR has been updated but the veracity of it's output cannot be spoken for. An expert should take a look to make sure the results are reasonable. Reviewers: rnk, JosephTremoulet, andrew.w.kaylor Differential Revision: http://reviews.llvm.org/D15139 llvm-svn: 255422
2015-12-12 13:38:55 +08:00
%clean.1 = cleanuppad within none []
[IR] Remove terminatepad It turns out that terminatepad gives little benefit over a cleanuppad which calls the termination function. This is not sufficient to implement fully generic filters but MSVC doesn't support them which makes terminatepad a little over-designed. Depends on D15478. Differential Revision: http://reviews.llvm.org/D15479 llvm-svn: 255522
2015-12-15 02:34:23 +08:00
unreachable
[IR] Reformulate LLVM's EH funclet IR While we have successfully implemented a funclet-oriented EH scheme on top of LLVM IR, our scheme has some notable deficiencies: - catchendpad and cleanupendpad are necessary in the current design but they are difficult to explain to others, even to seasoned LLVM experts. - catchendpad and cleanupendpad are optimization barriers. They cannot be split and force all potentially throwing call-sites to be invokes. This has a noticable effect on the quality of our code generation. - catchpad, while similar in some aspects to invoke, is fairly awkward. It is unsplittable, starts a funclet, and has control flow to other funclets. - The nesting relationship between funclets is currently a property of control flow edges. Because of this, we are forced to carefully analyze the flow graph to see if there might potentially exist illegal nesting among funclets. While we have logic to clone funclets when they are illegally nested, it would be nicer if we had a representation which forbade them upfront. Let's clean this up a bit by doing the following: - Instead, make catchpad more like cleanuppad and landingpad: no control flow, just a bunch of simple operands; catchpad would be splittable. - Introduce catchswitch, a control flow instruction designed to model the constraints of funclet oriented EH. - Make funclet scoping explicit by having funclet instructions consume the token produced by the funclet which contains them. - Remove catchendpad and cleanupendpad. Their presence can be inferred implicitly using coloring information. N.B. The state numbering code for the CLR has been updated but the veracity of it's output cannot be spoken for. An expert should take a look to make sure the results are reasonable. Reviewers: rnk, JosephTremoulet, andrew.w.kaylor Differential Revision: http://reviews.llvm.org/D15139 llvm-svn: 255422
2015-12-12 13:38:55 +08:00
; CHECK: %clean.1 = cleanuppad within none []
[IR] Remove terminatepad It turns out that terminatepad gives little benefit over a cleanuppad which calls the termination function. This is not sufficient to implement fully generic filters but MSVC doesn't support them which makes terminatepad a little over-designed. Depends on D15478. Differential Revision: http://reviews.llvm.org/D15479 llvm-svn: 255522
2015-12-15 02:34:23 +08:00
; CHECK-NEXT: unreachable
[Bitcode] Add compatibility tests for new instructions Adds basic compatibility tests for the following instructions: catchpad, catchendpad, cleanuppad, cleanupendpad, terminatepad, cleanupret, catchret llvm-svn: 247087
2015-09-09 06:33:23 +08:00
normal:
ret i32 0
}
[IR] Reformulate LLVM's EH funclet IR While we have successfully implemented a funclet-oriented EH scheme on top of LLVM IR, our scheme has some notable deficiencies: - catchendpad and cleanupendpad are necessary in the current design but they are difficult to explain to others, even to seasoned LLVM experts. - catchendpad and cleanupendpad are optimization barriers. They cannot be split and force all potentially throwing call-sites to be invokes. This has a noticable effect on the quality of our code generation. - catchpad, while similar in some aspects to invoke, is fairly awkward. It is unsplittable, starts a funclet, and has control flow to other funclets. - The nesting relationship between funclets is currently a property of control flow edges. Because of this, we are forced to carefully analyze the flow graph to see if there might potentially exist illegal nesting among funclets. While we have logic to clone funclets when they are illegally nested, it would be nicer if we had a representation which forbade them upfront. Let's clean this up a bit by doing the following: - Instead, make catchpad more like cleanuppad and landingpad: no control flow, just a bunch of simple operands; catchpad would be splittable. - Introduce catchswitch, a control flow instruction designed to model the constraints of funclet oriented EH. - Make funclet scoping explicit by having funclet instructions consume the token produced by the funclet which contains them. - Remove catchendpad and cleanupendpad. Their presence can be inferred implicitly using coloring information. N.B. The state numbering code for the CLR has been updated but the veracity of it's output cannot be spoken for. An expert should take a look to make sure the results are reasonable. Reviewers: rnk, JosephTremoulet, andrew.w.kaylor Differential Revision: http://reviews.llvm.org/D15139 llvm-svn: 255422
2015-12-12 13:38:55 +08:00
;
[Bitcode] Add compatibility tests for new instructions Adds basic compatibility tests for the following instructions: catchpad, catchendpad, cleanuppad, cleanupendpad, terminatepad, cleanupret, catchret llvm-svn: 247087
2015-09-09 06:33:23 +08:00
define i32 @instructions.win_eh.2() personality i32 -4 {
entry:
[IR] Reformulate LLVM's EH funclet IR While we have successfully implemented a funclet-oriented EH scheme on top of LLVM IR, our scheme has some notable deficiencies: - catchendpad and cleanupendpad are necessary in the current design but they are difficult to explain to others, even to seasoned LLVM experts. - catchendpad and cleanupendpad are optimization barriers. They cannot be split and force all potentially throwing call-sites to be invokes. This has a noticable effect on the quality of our code generation. - catchpad, while similar in some aspects to invoke, is fairly awkward. It is unsplittable, starts a funclet, and has control flow to other funclets. - The nesting relationship between funclets is currently a property of control flow edges. Because of this, we are forced to carefully analyze the flow graph to see if there might potentially exist illegal nesting among funclets. While we have logic to clone funclets when they are illegally nested, it would be nicer if we had a representation which forbade them upfront. Let's clean this up a bit by doing the following: - Instead, make catchpad more like cleanuppad and landingpad: no control flow, just a bunch of simple operands; catchpad would be splittable. - Introduce catchswitch, a control flow instruction designed to model the constraints of funclet oriented EH. - Make funclet scoping explicit by having funclet instructions consume the token produced by the funclet which contains them. - Remove catchendpad and cleanupendpad. Their presence can be inferred implicitly using coloring information. N.B. The state numbering code for the CLR has been updated but the veracity of it's output cannot be spoken for. An expert should take a look to make sure the results are reasonable. Reviewers: rnk, JosephTremoulet, andrew.w.kaylor Differential Revision: http://reviews.llvm.org/D15139 llvm-svn: 255422
2015-12-12 13:38:55 +08:00
invoke void @f.ccc() to label %invoke.cont unwind label %catchswitch
[Bitcode] Add compatibility tests for new instructions Adds basic compatibility tests for the following instructions: catchpad, catchendpad, cleanuppad, cleanupendpad, terminatepad, cleanupret, catchret llvm-svn: 247087
2015-09-09 06:33:23 +08:00
invoke.cont:
invoke void @f.ccc() to label %continue unwind label %cleanup
cleanup:
[IR] Reformulate LLVM's EH funclet IR While we have successfully implemented a funclet-oriented EH scheme on top of LLVM IR, our scheme has some notable deficiencies: - catchendpad and cleanupendpad are necessary in the current design but they are difficult to explain to others, even to seasoned LLVM experts. - catchendpad and cleanupendpad are optimization barriers. They cannot be split and force all potentially throwing call-sites to be invokes. This has a noticable effect on the quality of our code generation. - catchpad, while similar in some aspects to invoke, is fairly awkward. It is unsplittable, starts a funclet, and has control flow to other funclets. - The nesting relationship between funclets is currently a property of control flow edges. Because of this, we are forced to carefully analyze the flow graph to see if there might potentially exist illegal nesting among funclets. While we have logic to clone funclets when they are illegally nested, it would be nicer if we had a representation which forbade them upfront. Let's clean this up a bit by doing the following: - Instead, make catchpad more like cleanuppad and landingpad: no control flow, just a bunch of simple operands; catchpad would be splittable. - Introduce catchswitch, a control flow instruction designed to model the constraints of funclet oriented EH. - Make funclet scoping explicit by having funclet instructions consume the token produced by the funclet which contains them. - Remove catchendpad and cleanupendpad. Their presence can be inferred implicitly using coloring information. N.B. The state numbering code for the CLR has been updated but the veracity of it's output cannot be spoken for. An expert should take a look to make sure the results are reasonable. Reviewers: rnk, JosephTremoulet, andrew.w.kaylor Differential Revision: http://reviews.llvm.org/D15139 llvm-svn: 255422
2015-12-12 13:38:55 +08:00
%clean = cleanuppad within none []
; CHECK: %clean = cleanuppad within none []
cleanupret from %clean unwind to caller
; CHECK: cleanupret from %clean unwind to caller
catchswitch:
%cs = catchswitch within none [label %catchpad] unwind label %terminate
[Bitcode] Add compatibility tests for new instructions Adds basic compatibility tests for the following instructions: catchpad, catchendpad, cleanuppad, cleanupendpad, terminatepad, cleanupret, catchret llvm-svn: 247087
2015-09-09 06:33:23 +08:00
catchpad:
[IR] Reformulate LLVM's EH funclet IR While we have successfully implemented a funclet-oriented EH scheme on top of LLVM IR, our scheme has some notable deficiencies: - catchendpad and cleanupendpad are necessary in the current design but they are difficult to explain to others, even to seasoned LLVM experts. - catchendpad and cleanupendpad are optimization barriers. They cannot be split and force all potentially throwing call-sites to be invokes. This has a noticable effect on the quality of our code generation. - catchpad, while similar in some aspects to invoke, is fairly awkward. It is unsplittable, starts a funclet, and has control flow to other funclets. - The nesting relationship between funclets is currently a property of control flow edges. Because of this, we are forced to carefully analyze the flow graph to see if there might potentially exist illegal nesting among funclets. While we have logic to clone funclets when they are illegally nested, it would be nicer if we had a representation which forbade them upfront. Let's clean this up a bit by doing the following: - Instead, make catchpad more like cleanuppad and landingpad: no control flow, just a bunch of simple operands; catchpad would be splittable. - Introduce catchswitch, a control flow instruction designed to model the constraints of funclet oriented EH. - Make funclet scoping explicit by having funclet instructions consume the token produced by the funclet which contains them. - Remove catchendpad and cleanupendpad. Their presence can be inferred implicitly using coloring information. N.B. The state numbering code for the CLR has been updated but the veracity of it's output cannot be spoken for. An expert should take a look to make sure the results are reasonable. Reviewers: rnk, JosephTremoulet, andrew.w.kaylor Differential Revision: http://reviews.llvm.org/D15139 llvm-svn: 255422
2015-12-12 13:38:55 +08:00
%catch = catchpad within %cs []
br label %body
; CHECK: %catch = catchpad within %cs []
; CHECK-NEXT: br label %body
[Bitcode] Add compatibility tests for new instructions Adds basic compatibility tests for the following instructions: catchpad, catchendpad, cleanuppad, cleanupendpad, terminatepad, cleanupret, catchret llvm-svn: 247087
2015-09-09 06:33:23 +08:00
body:
[WinEH] Verify unwind edges against EH pad tree Summary: Funclet EH personalities require a tree-like nesting among funclets (enforced by the ParentPad linkage in the IR), and also require that unwind edges conform to certain rules with respect to the tree: - An unwind edge may exit 0 or more ancestor pads - An unwind edge must enter exactly one EH pad, which must be distinct from any exited pads - A cleanupret's edge must exit its cleanuppad Describe these rules in the LangRef, and enforce them in the verifier. Reviewers: rnk, majnemer, andrew.w.kaylor Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D15961 llvm-svn: 257272
2016-01-10 12:28:38 +08:00
invoke void @f.ccc() [ "funclet"(token %catch) ]
to label %continue unwind label %terminate.inner
[IR] Reformulate LLVM's EH funclet IR While we have successfully implemented a funclet-oriented EH scheme on top of LLVM IR, our scheme has some notable deficiencies: - catchendpad and cleanupendpad are necessary in the current design but they are difficult to explain to others, even to seasoned LLVM experts. - catchendpad and cleanupendpad are optimization barriers. They cannot be split and force all potentially throwing call-sites to be invokes. This has a noticable effect on the quality of our code generation. - catchpad, while similar in some aspects to invoke, is fairly awkward. It is unsplittable, starts a funclet, and has control flow to other funclets. - The nesting relationship between funclets is currently a property of control flow edges. Because of this, we are forced to carefully analyze the flow graph to see if there might potentially exist illegal nesting among funclets. While we have logic to clone funclets when they are illegally nested, it would be nicer if we had a representation which forbade them upfront. Let's clean this up a bit by doing the following: - Instead, make catchpad more like cleanuppad and landingpad: no control flow, just a bunch of simple operands; catchpad would be splittable. - Introduce catchswitch, a control flow instruction designed to model the constraints of funclet oriented EH. - Make funclet scoping explicit by having funclet instructions consume the token produced by the funclet which contains them. - Remove catchendpad and cleanupendpad. Their presence can be inferred implicitly using coloring information. N.B. The state numbering code for the CLR has been updated but the veracity of it's output cannot be spoken for. An expert should take a look to make sure the results are reasonable. Reviewers: rnk, JosephTremoulet, andrew.w.kaylor Differential Revision: http://reviews.llvm.org/D15139 llvm-svn: 255422
2015-12-12 13:38:55 +08:00
catchret from %catch to label %return
; CHECK: catchret from %catch to label %return
[Bitcode] Add compatibility tests for new instructions Adds basic compatibility tests for the following instructions: catchpad, catchendpad, cleanuppad, cleanupendpad, terminatepad, cleanupret, catchret llvm-svn: 247087
2015-09-09 06:33:23 +08:00
return:
ret i32 0
[WinEH] Tighten parentPad verifier checks Summary: A catchswitch cannot be a parent of a cleanuppad or another catchswitch. Reviewers: rnk, andrew.w.kaylor, majnemer Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D15841 llvm-svn: 256690
2016-01-02 23:24:24 +08:00
terminate.inner:
cleanuppad within %catch []
unreachable
; CHECK: cleanuppad within %catch []
; CHECK-NEXT: unreachable
[Bitcode] Add compatibility tests for new instructions Adds basic compatibility tests for the following instructions: catchpad, catchendpad, cleanuppad, cleanupendpad, terminatepad, cleanupret, catchret llvm-svn: 247087
2015-09-09 06:33:23 +08:00
terminate:
[WinEH] Tighten parentPad verifier checks Summary: A catchswitch cannot be a parent of a cleanuppad or another catchswitch. Reviewers: rnk, andrew.w.kaylor, majnemer Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D15841 llvm-svn: 256690
2016-01-02 23:24:24 +08:00
cleanuppad within none []
[IR] Remove terminatepad It turns out that terminatepad gives little benefit over a cleanuppad which calls the termination function. This is not sufficient to implement fully generic filters but MSVC doesn't support them which makes terminatepad a little over-designed. Depends on D15478. Differential Revision: http://reviews.llvm.org/D15479 llvm-svn: 255522
2015-12-15 02:34:23 +08:00
unreachable
[WinEH] Tighten parentPad verifier checks Summary: A catchswitch cannot be a parent of a cleanuppad or another catchswitch. Reviewers: rnk, andrew.w.kaylor, majnemer Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D15841 llvm-svn: 256690
2016-01-02 23:24:24 +08:00
; CHECK: cleanuppad within none []
[IR] Remove terminatepad It turns out that terminatepad gives little benefit over a cleanuppad which calls the termination function. This is not sufficient to implement fully generic filters but MSVC doesn't support them which makes terminatepad a little over-designed. Depends on D15478. Differential Revision: http://reviews.llvm.org/D15479 llvm-svn: 255522
2015-12-15 02:34:23 +08:00
; CHECK-NEXT: unreachable
[Bitcode] Add compatibility tests for new instructions Adds basic compatibility tests for the following instructions: catchpad, catchendpad, cleanuppad, cleanupendpad, terminatepad, cleanupret, catchret llvm-svn: 247087
2015-09-09 06:33:23 +08:00
continue:
ret i32 0
}
[IR] Add a dedicated FNeg IR Instruction The IEEE-754 Standard makes it clear that fneg(x) and fsub(-0.0, x) are two different operations. The former is a bitwise operation, while the latter is an arithmetic operation. This patch creates a dedicated FNeg IR Instruction to model that behavior. Differential Revision: https://reviews.llvm.org/D53877 llvm-svn: 346774
2018-11-14 02:15:47 +08:00
; Instructions -- Unary Operations
[IR] Add Freeze instruction Summary: - Define Instruction::Freeze, let it be UnaryOperator - Add support for freeze to LLLexer/LLParser/BitcodeReader/BitcodeWriter The format is `%x = freeze <ty> %v` - Add support for freeze instruction to llvm-c interface. - Add m_Freeze in PatternMatch. - Erase freeze when lowering IR to SelDag. Reviewers: deadalnix, hfinkel, efriedma, lebedev.ri, nlopes, jdoerfert, regehr, filcab, delcypher, whitequark Reviewed By: lebedev.ri, jdoerfert Subscribers: jfb, kristof.beyls, hiraditya, lebedev.ri, steven_wu, dexonsmith, xbolva00, delcypher, spatel, regehr, trentxintong, vsk, filcab, nlopes, mehdi_amini, deadalnix, llvm-commits Differential Revision: https://reviews.llvm.org/D29011
2019-11-05 14:53:22 +08:00
define void @instructions.unops(double %op1, i32 %op2, <2 x i32> %op3, i8* %op4) {
[IR] Add a dedicated FNeg IR Instruction The IEEE-754 Standard makes it clear that fneg(x) and fsub(-0.0, x) are two different operations. The former is a bitwise operation, while the latter is an arithmetic operation. This patch creates a dedicated FNeg IR Instruction to model that behavior. Differential Revision: https://reviews.llvm.org/D53877 llvm-svn: 346774
2018-11-14 02:15:47 +08:00
fneg double %op1
; CHECK: fneg double %op1
[IR] Add Freeze instruction Summary: - Define Instruction::Freeze, let it be UnaryOperator - Add support for freeze to LLLexer/LLParser/BitcodeReader/BitcodeWriter The format is `%x = freeze <ty> %v` - Add support for freeze instruction to llvm-c interface. - Add m_Freeze in PatternMatch. - Erase freeze when lowering IR to SelDag. Reviewers: deadalnix, hfinkel, efriedma, lebedev.ri, nlopes, jdoerfert, regehr, filcab, delcypher, whitequark Reviewed By: lebedev.ri, jdoerfert Subscribers: jfb, kristof.beyls, hiraditya, lebedev.ri, steven_wu, dexonsmith, xbolva00, delcypher, spatel, regehr, trentxintong, vsk, filcab, nlopes, mehdi_amini, deadalnix, llvm-commits Differential Revision: https://reviews.llvm.org/D29011
2019-11-05 14:53:22 +08:00
freeze i32 %op2
; CHECK: freeze i32 %op2
freeze double %op1
; CHECK: freeze double %op1
freeze <2 x i32> %op3
; CHECK: freeze <2 x i32> %op3
freeze i8* %op4
; CHECK: freeze i8* %op4
[IR] Add a dedicated FNeg IR Instruction The IEEE-754 Standard makes it clear that fneg(x) and fsub(-0.0, x) are two different operations. The former is a bitwise operation, while the latter is an arithmetic operation. This patch creates a dedicated FNeg IR Instruction to model that behavior. Differential Revision: https://reviews.llvm.org/D53877 llvm-svn: 346774
2018-11-14 02:15:47 +08:00
ret void
}
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
; Instructions -- Binary Operations
define void @instructions.binops(i8 %op1, i8 %op2) {
; nuw x nsw
add i8 %op1, %op2
; CHECK: add i8 %op1, %op2
add nuw i8 %op1, %op2
; CHECK: add nuw i8 %op1, %op2
add nsw i8 %op1, %op2
; CHECK: add nsw i8 %op1, %op2
add nuw nsw i8 %op1, %op2
; CHECK: add nuw nsw i8 %op1, %op2
sub i8 %op1, %op2
; CHECK: sub i8 %op1, %op2
sub nuw i8 %op1, %op2
; CHECK: sub nuw i8 %op1, %op2
sub nsw i8 %op1, %op2
; CHECK: sub nsw i8 %op1, %op2
sub nuw nsw i8 %op1, %op2
; CHECK: sub nuw nsw i8 %op1, %op2
mul i8 %op1, %op2
; CHECK: mul i8 %op1, %op2
mul nuw i8 %op1, %op2
; CHECK: mul nuw i8 %op1, %op2
mul nsw i8 %op1, %op2
; CHECK: mul nsw i8 %op1, %op2
mul nuw nsw i8 %op1, %op2
; CHECK: mul nuw nsw i8 %op1, %op2
; exact
udiv i8 %op1, %op2
; CHECK: udiv i8 %op1, %op2
udiv exact i8 %op1, %op2
; CHECK: udiv exact i8 %op1, %op2
sdiv i8 %op1, %op2
; CHECK: sdiv i8 %op1, %op2
sdiv exact i8 %op1, %op2
; CHECK: sdiv exact i8 %op1, %op2
; none
urem i8 %op1, %op2
; CHECK: urem i8 %op1, %op2
srem i8 %op1, %op2
; CHECK: srem i8 %op1, %op2
ret void
}
; Instructions -- Bitwise Binary Operations
define void @instructions.bitwise_binops(i8 %op1, i8 %op2) {
; nuw x nsw
shl i8 %op1, %op2
; CHECK: shl i8 %op1, %op2
shl nuw i8 %op1, %op2
; CHECK: shl nuw i8 %op1, %op2
shl nsw i8 %op1, %op2
; CHECK: shl nsw i8 %op1, %op2
shl nuw nsw i8 %op1, %op2
; CHECK: shl nuw nsw i8 %op1, %op2
; exact
lshr i8 %op1, %op2
; CHECK: lshr i8 %op1, %op2
lshr exact i8 %op1, %op2
; CHECK: lshr exact i8 %op1, %op2
ashr i8 %op1, %op2
; CHECK: ashr i8 %op1, %op2
ashr exact i8 %op1, %op2
; CHECK: ashr exact i8 %op1, %op2
; none
and i8 %op1, %op2
; CHECK: and i8 %op1, %op2
or i8 %op1, %op2
; CHECK: or i8 %op1, %op2
xor i8 %op1, %op2
; CHECK: xor i8 %op1, %op2
ret void
}
; Instructions -- Vector Operations
define void @instructions.vectorops(<4 x float> %vec, <4 x float> %vec2) {
extractelement <4 x float> %vec, i8 0
; CHECK: extractelement <4 x float> %vec, i8 0
insertelement <4 x float> %vec, float 3.500000e+00, i8 0
; CHECK: insertelement <4 x float> %vec, float 3.500000e+00, i8 0
shufflevector <4 x float> %vec, <4 x float> %vec2, <2 x i32> zeroinitializer
; CHECK: shufflevector <4 x float> %vec, <4 x float> %vec2, <2 x i32> zeroinitializer
ret void
}
; Instructions -- Aggregate Operations
define void @instructions.aggregateops({ i8, i32 } %up, <{ i8, i32 }> %p,
[3 x i8] %arr, { i8, { i32 }} %n,
<2 x i8*> %pvec, <2 x i64> %offsets) {
extractvalue { i8, i32 } %up, 0
; CHECK: extractvalue { i8, i32 } %up, 0
extractvalue <{ i8, i32 }> %p, 1
; CHECK: extractvalue <{ i8, i32 }> %p, 1
extractvalue [3 x i8] %arr, 2
; CHECK: extractvalue [3 x i8] %arr, 2
extractvalue { i8, { i32 } } %n, 1, 0
; CHECK: extractvalue { i8, { i32 } } %n, 1, 0
insertvalue { i8, i32 } %up, i8 1, 0
; CHECK: insertvalue { i8, i32 } %up, i8 1, 0
insertvalue <{ i8, i32 }> %p, i32 2, 1
; CHECK: insertvalue <{ i8, i32 }> %p, i32 2, 1
insertvalue [3 x i8] %arr, i8 0, 0
; CHECK: insertvalue [3 x i8] %arr, i8 0, 0
insertvalue { i8, { i32 } } %n, i32 0, 1, 0
; CHECK: insertvalue { i8, { i32 } } %n, i32 0, 1, 0
%up.ptr = alloca { i8, i32 }
%p.ptr = alloca <{ i8, i32 }>
%arr.ptr = alloca [3 x i8]
%n.ptr = alloca { i8, { i32 } }
getelementptr { i8, i32 }, { i8, i32 }* %up.ptr, i8 0
; CHECK: getelementptr { i8, i32 }, { i8, i32 }* %up.ptr, i8 0
getelementptr <{ i8, i32 }>, <{ i8, i32 }>* %p.ptr, i8 1
; CHECK: getelementptr <{ i8, i32 }>, <{ i8, i32 }>* %p.ptr, i8 1
getelementptr [3 x i8], [3 x i8]* %arr.ptr, i8 2
; CHECK: getelementptr [3 x i8], [3 x i8]* %arr.ptr, i8 2
getelementptr { i8, { i32 } }, { i8, { i32 } }* %n.ptr, i32 0, i32 1
; CHECK: getelementptr { i8, { i32 } }, { i8, { i32 } }* %n.ptr, i32 0, i32 1
getelementptr inbounds { i8, { i32 } }, { i8, { i32 } }* %n.ptr, i32 1, i32 0
; CHECK: getelementptr inbounds { i8, { i32 } }, { i8, { i32 } }* %n.ptr, i32 1, i32 0
getelementptr i8, <2 x i8*> %pvec, <2 x i64> %offsets
; CHECK: getelementptr i8, <2 x i8*> %pvec, <2 x i64> %offsets
ret void
}
; Instructions -- Memory Access and Addressing Operations
!7 = !{i32 1}
!8 = !{}
!9 = !{i64 4}
define void @instructions.memops(i32** %base) {
alloca i32, i8 4, align 4
; CHECK: alloca i32, i8 4, align 4
alloca inalloca i32, i8 4, align 4
; CHECK: alloca inalloca i32, i8 4, align 4
load i32*, i32** %base, align 8, !invariant.load !7, !nontemporal !8, !nonnull !7, !dereferenceable !9, !dereferenceable_or_null !9
; CHECK: load i32*, i32** %base, align 8, !invariant.load !7, !nontemporal !8, !nonnull !7, !dereferenceable !9, !dereferenceable_or_null !9
load volatile i32*, i32** %base, align 8, !invariant.load !7, !nontemporal !8, !nonnull !7, !dereferenceable !9, !dereferenceable_or_null !9
; CHECK: load volatile i32*, i32** %base, align 8, !invariant.load !7, !nontemporal !8, !nonnull !7, !dereferenceable !9, !dereferenceable_or_null !9
store i32* null, i32** %base, align 4, !nontemporal !8
; CHECK: store i32* null, i32** %base, align 4, !nontemporal !8
store volatile i32* null, i32** %base, align 4, !nontemporal !8
; CHECK: store volatile i32* null, i32** %base, align 4, !nontemporal !8
ret void
}
; Instructions -- Conversion Operations
define void @instructions.conversions() {
trunc i32 -1 to i1
; CHECK: trunc i32 -1 to i1
zext i32 -1 to i64
; CHECK: zext i32 -1 to i64
sext i32 -1 to i64
; CHECK: sext i32 -1 to i64
fptrunc float undef to half
; CHECK: fptrunc float undef to half
fpext half undef to float
; CHECK: fpext half undef to float
fptoui float undef to i32
; CHECK: fptoui float undef to i32
fptosi float undef to i32
; CHECK: fptosi float undef to i32
uitofp i32 1 to float
; CHECK: uitofp i32 1 to float
sitofp i32 -1 to float
; CHECK: sitofp i32 -1 to float
ptrtoint i8* null to i64
; CHECK: ptrtoint i8* null to i64
inttoptr i64 0 to i8*
; CHECK: inttoptr i64 0 to i8*
bitcast i32 0 to i32
; CHECK: bitcast i32 0 to i32
addrspacecast i32* null to i32 addrspace(1)*
; CHECK: addrspacecast i32* null to i32 addrspace(1)*
ret void
}
; Instructions -- Other Operations
define void @instructions.other(i32 %op1, i32 %op2, half %fop1, half %fop2) {
entry:
icmp eq i32 %op1, %op2
; CHECK: icmp eq i32 %op1, %op2
icmp ne i32 %op1, %op2
; CHECK: icmp ne i32 %op1, %op2
icmp ugt i32 %op1, %op2
; CHECK: icmp ugt i32 %op1, %op2
icmp uge i32 %op1, %op2
; CHECK: icmp uge i32 %op1, %op2
icmp ult i32 %op1, %op2
; CHECK: icmp ult i32 %op1, %op2
icmp ule i32 %op1, %op2
; CHECK: icmp ule i32 %op1, %op2
icmp sgt i32 %op1, %op2
; CHECK: icmp sgt i32 %op1, %op2
icmp sge i32 %op1, %op2
; CHECK: icmp sge i32 %op1, %op2
icmp slt i32 %op1, %op2
; CHECK: icmp slt i32 %op1, %op2
icmp sle i32 %op1, %op2
; CHECK: icmp sle i32 %op1, %op2
fcmp false half %fop1, %fop2
; CHECK: fcmp false half %fop1, %fop2
fcmp oeq half %fop1, %fop2
; CHECK: fcmp oeq half %fop1, %fop2
fcmp ogt half %fop1, %fop2
; CHECK: fcmp ogt half %fop1, %fop2
fcmp oge half %fop1, %fop2
; CHECK: fcmp oge half %fop1, %fop2
fcmp olt half %fop1, %fop2
; CHECK: fcmp olt half %fop1, %fop2
fcmp ole half %fop1, %fop2
; CHECK: fcmp ole half %fop1, %fop2
fcmp one half %fop1, %fop2
; CHECK: fcmp one half %fop1, %fop2
fcmp ord half %fop1, %fop2
; CHECK: fcmp ord half %fop1, %fop2
fcmp ueq half %fop1, %fop2
; CHECK: fcmp ueq half %fop1, %fop2
fcmp ugt half %fop1, %fop2
; CHECK: fcmp ugt half %fop1, %fop2
fcmp uge half %fop1, %fop2
; CHECK: fcmp uge half %fop1, %fop2
fcmp ult half %fop1, %fop2
; CHECK: fcmp ult half %fop1, %fop2
fcmp ule half %fop1, %fop2
; CHECK: fcmp ule half %fop1, %fop2
fcmp une half %fop1, %fop2
; CHECK: fcmp une half %fop1, %fop2
fcmp uno half %fop1, %fop2
; CHECK: fcmp uno half %fop1, %fop2
fcmp true half %fop1, %fop2
; CHECK: fcmp true half %fop1, %fop2
br label %exit
L1:
%v1 = add i32 %op1, %op2
br label %exit
L2:
%v2 = add i32 %op1, %op2
br label %exit
exit:
phi i32 [ %v1, %L1 ], [ %v2, %L2 ], [ %op1, %entry ]
; CHECK: phi i32 [ %v1, %L1 ], [ %v2, %L2 ], [ %op1, %entry ]
select i1 true, i32 0, i32 1
; CHECK: select i1 true, i32 0, i32 1
select <2 x i1> <i1 true, i1 false>, <2 x i8> <i8 2, i8 3>, <2 x i8> <i8 3, i8 2>
; CHECK: select <2 x i1> <i1 true, i1 false>, <2 x i8> <i8 2, i8 3>, <2 x i8> <i8 3, i8 2>
call void @f.nobuiltin() builtin
Minor movement of one function with now-correct strictfp attribute to pass against the latest version of D68233.
2019-10-24 00:12:56 +08:00
; CHECK: call void @f.nobuiltin() #44
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
call fastcc noalias i32* @f.noalias() noinline
; CHECK: call fastcc noalias i32* @f.noalias() #12
tail call ghccc nonnull i32* @f.nonnull() minsize
; CHECK: tail call ghccc nonnull i32* @f.nonnull() #7
ret void
}
define void @instructions.call_musttail(i8* inalloca %val) {
musttail call void @f.param.inalloca(i8* inalloca %val)
; CHECK: musttail call void @f.param.inalloca(i8* inalloca %val)
ret void
}
Add 'notail' marker for call instructions. This marker prevents optimization passes from adding 'tail' or 'musttail' markers to a call. Is is used to prevent tail call optimization from being performed on the call. rdar://problem/22667622 Differential Revision: http://reviews.llvm.org/D12923 llvm-svn: 252368
2015-11-07 07:55:38 +08:00
define void @instructions.call_notail() {
notail call void @f1()
; CHECK: notail call void @f1()
ret void
}
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
define void @instructions.landingpad() personality i32 -2 {
invoke void @llvm.donothing() to label %proceed unwind label %catch1
invoke void @llvm.donothing() to label %proceed unwind label %catch2
invoke void @llvm.donothing() to label %proceed unwind label %catch3
invoke void @llvm.donothing() to label %proceed unwind label %catch4
catch1:
landingpad i32
; CHECK: landingpad i32
cleanup
; CHECK: cleanup
br label %proceed
catch2:
landingpad i32
; CHECK: landingpad i32
cleanup
; CHECK: cleanup
catch i32* null
; CHECK: catch i32* null
br label %proceed
catch3:
landingpad i32
; CHECK: landingpad i32
cleanup
; CHECK: cleanup
catch i32* null
; CHECK: catch i32* null
catch i32* null
; CHECK: catch i32* null
br label %proceed
catch4:
landingpad i32
; CHECK: landingpad i32
filter [2 x i32] zeroinitializer
; CHECK: filter [2 x i32] zeroinitializer
br label %proceed
proceed:
ret void
}
;; Intrinsic Functions
; Intrinsic Functions -- Variable Argument Handling
declare void @llvm.va_start(i8*)
declare void @llvm.va_copy(i8*, i8*)
declare void @llvm.va_end(i8*)
define void @instructions.va_arg(i8* %v, ...) {
%ap = alloca i8*
%ap2 = bitcast i8** %ap to i8*
call void @llvm.va_start(i8* %ap2)
; CHECK: call void @llvm.va_start(i8* %ap2)
va_arg i8* %ap2, i32
; CHECK: va_arg i8* %ap2, i32
call void @llvm.va_copy(i8* %v, i8* %ap2)
; CHECK: call void @llvm.va_copy(i8* %v, i8* %ap2)
call void @llvm.va_end(i8* %ap2)
; CHECK: call void @llvm.va_end(i8* %ap2)
ret void
}
; Intrinsic Functions -- Accurate Garbage Collection
declare void @llvm.gcroot(i8**, i8*)
declare i8* @llvm.gcread(i8*, i8**)
declare void @llvm.gcwrite(i8*, i8*, i8**)
define void @intrinsics.gc() gc "shadow-stack" {
%ptrloc = alloca i8*
call void @llvm.gcroot(i8** %ptrloc, i8* null)
; CHECK: call void @llvm.gcroot(i8** %ptrloc, i8* null)
call i8* @llvm.gcread(i8* null, i8** %ptrloc)
; CHECK: call i8* @llvm.gcread(i8* null, i8** %ptrloc)
%ref = alloca i8
call void @llvm.gcwrite(i8* %ref, i8* null, i8** %ptrloc)
; CHECK: call void @llvm.gcwrite(i8* %ref, i8* null, i8** %ptrloc)
ret void
}
; Intrinsic Functions -- Code Generation
declare i8* @llvm.returnaddress(i32)
declare i8* @llvm.frameaddress(i32)
declare i32 @llvm.read_register.i32(metadata)
declare i64 @llvm.read_register.i64(metadata)
declare void @llvm.write_register.i32(metadata, i32)
declare void @llvm.write_register.i64(metadata, i64)
declare i8* @llvm.stacksave()
declare void @llvm.stackrestore(i8*)
Allow prefetching from non-zero address spaces Summary: This is useful for targets which have prefetch instructions for non-default address spaces. <rdar://problem/42662136> Subscribers: nemanjai, javed.absar, hiraditya, kbarton, jkorous, dexonsmith, cfe-commits, llvm-commits, RKSimon, hfinkel, t.p.northover, craig.topper, anemet Tags: #clang, #llvm Differential Revision: https://reviews.llvm.org/D65254 llvm-svn: 367032
2019-07-26 00:11:57 +08:00
declare void @llvm.prefetch.p0i8(i8*, i32, i32, i32)
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
declare void @llvm.pcmarker(i32)
declare i64 @llvm.readcyclecounter()
declare void @llvm.clear_cache(i8*, i8*)
declare void @llvm.instrprof_increment(i8*, i64, i32, i32)
!10 = !{!"rax"}
define void @intrinsics.codegen() {
call i8* @llvm.returnaddress(i32 1)
; CHECK: call i8* @llvm.returnaddress(i32 1)
call i8* @llvm.frameaddress(i32 1)
Added address-space mangling for stack related intrinsics Modified the following 3 intrinsics: int_addressofreturnaddress, int_frameaddress & int_sponentry. Reviewed By: arsenm Differential Revision: https://reviews.llvm.org/D64561 llvm-svn: 366679
2019-07-22 20:42:48 +08:00
; CHECK: call i8* @llvm.frameaddress.p0i8(i32 1)
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
call i32 @llvm.read_register.i32(metadata !10)
; CHECK: call i32 @llvm.read_register.i32(metadata !10)
call i64 @llvm.read_register.i64(metadata !10)
; CHECK: call i64 @llvm.read_register.i64(metadata !10)
call void @llvm.write_register.i32(metadata !10, i32 0)
; CHECK: call void @llvm.write_register.i32(metadata !10, i32 0)
call void @llvm.write_register.i64(metadata !10, i64 0)
; CHECK: call void @llvm.write_register.i64(metadata !10, i64 0)
%stack = call i8* @llvm.stacksave()
; CHECK: %stack = call i8* @llvm.stacksave()
call void @llvm.stackrestore(i8* %stack)
; CHECK: call void @llvm.stackrestore(i8* %stack)
Allow prefetching from non-zero address spaces Summary: This is useful for targets which have prefetch instructions for non-default address spaces. <rdar://problem/42662136> Subscribers: nemanjai, javed.absar, hiraditya, kbarton, jkorous, dexonsmith, cfe-commits, llvm-commits, RKSimon, hfinkel, t.p.northover, craig.topper, anemet Tags: #clang, #llvm Differential Revision: https://reviews.llvm.org/D65254 llvm-svn: 367032
2019-07-26 00:11:57 +08:00
call void @llvm.prefetch.p0i8(i8* %stack, i32 0, i32 3, i32 0)
; CHECK: call void @llvm.prefetch.p0i8(i8* %stack, i32 0, i32 3, i32 0)
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
call void @llvm.pcmarker(i32 1)
; CHECK: call void @llvm.pcmarker(i32 1)
call i64 @llvm.readcyclecounter()
; CHECK: call i64 @llvm.readcyclecounter()
call void @llvm.clear_cache(i8* null, i8* null)
; CHECK: call void @llvm.clear_cache(i8* null, i8* null)
call void @llvm.instrprof_increment(i8* null, i64 0, i32 0, i32 0)
; CHECK: call void @llvm.instrprof_increment(i8* null, i64 0, i32 0, i32 0)
ret void
}
declare void @llvm.localescape(...)
declare i8* @llvm.localrecover(i8* %func, i8* %fp, i32 %idx)
define void @intrinsics.localescape() {
%static.alloca = alloca i32
call void (...) @llvm.localescape(i32* %static.alloca)
; CHECK: call void (...) @llvm.localescape(i32* %static.alloca)
call void @intrinsics.localrecover()
ret void
}
define void @intrinsics.localrecover() {
%func = bitcast void ()* @intrinsics.localescape to i8*
%fp = call i8* @llvm.frameaddress(i32 1)
call i8* @llvm.localrecover(i8* %func, i8* %fp, i32 0)
; CHECK: call i8* @llvm.localrecover(i8* %func, i8* %fp, i32 0)
ret void
}
; We need this function to provide `uses' for some metadata tests.
define void @misc.metadata() {
call void @f1(), !srcloc !11
call void @f1(), !srcloc !12
call void @f1(), !srcloc !13
call void @f1(), !srcloc !14
ret void
}
[Bitcode][Asm] Teach LLVM to read and write operand bundles. Summary: This also adds the first set of tests for operand bundles. The optimizer has not been audited to ensure that it does the right thing with operand bundles. Depends on D12456. Reviewers: reames, chandlerc, majnemer, dexonsmith, kmod, JosephTremoulet, rnk, bogner Subscribers: maksfb, llvm-commits Differential Revision: http://reviews.llvm.org/D12457 llvm-svn: 248551
2015-09-25 07:34:52 +08:00
declare void @op_bundle_callee_0()
declare void @op_bundle_callee_1(i32,i32)
define void @call_with_operand_bundle0(i32* %ptr) {
; CHECK-LABEL: call_with_operand_bundle0(
entry:
%l = load i32, i32* %ptr
%x = add i32 42, 1
call void @op_bundle_callee_0() [ "foo"(i32 42, i64 100, i32 %x), "bar"(float 0.000000e+00, i64 100, i32 %l) ]
; CHECK: call void @op_bundle_callee_0() [ "foo"(i32 42, i64 100, i32 %x), "bar"(float 0.000000e+00, i64 100, i32 %l) ]
ret void
}
define void @call_with_operand_bundle1(i32* %ptr) {
; CHECK-LABEL: call_with_operand_bundle1(
entry:
%l = load i32, i32* %ptr
%x = add i32 42, 1
call void @op_bundle_callee_0()
call void @op_bundle_callee_0() [ "foo"() ]
call void @op_bundle_callee_0() [ "foo"(i32 42, i64 100, i32 %x), "bar"(float 0.000000e+00, i64 100, i32 %l) ]
; CHECK: @op_bundle_callee_0(){{$}}
; CHECK-NEXT: call void @op_bundle_callee_0() [ "foo"() ]
; CHECK-NEXT: call void @op_bundle_callee_0() [ "foo"(i32 42, i64 100, i32 %x), "bar"(float 0.000000e+00, i64 100, i32 %l) ]
ret void
}
define void @call_with_operand_bundle2(i32* %ptr) {
; CHECK-LABEL: call_with_operand_bundle2(
entry:
call void @op_bundle_callee_0() [ "foo"() ]
; CHECK: call void @op_bundle_callee_0() [ "foo"() ]
ret void
}
define void @call_with_operand_bundle3(i32* %ptr) {
; CHECK-LABEL: call_with_operand_bundle3(
entry:
%l = load i32, i32* %ptr
%x = add i32 42, 1
call void @op_bundle_callee_0() [ "foo"(i32 42, i64 100, i32 %x), "foo"(i32 42, float 0.000000e+00, i32 %l) ]
; CHECK: call void @op_bundle_callee_0() [ "foo"(i32 42, i64 100, i32 %x), "foo"(i32 42, float 0.000000e+00, i32 %l) ]
ret void
}
define void @call_with_operand_bundle4(i32* %ptr) {
; CHECK-LABEL: call_with_operand_bundle4(
entry:
%l = load i32, i32* %ptr
%x = add i32 42, 1
call void @op_bundle_callee_1(i32 10, i32 %x) [ "foo"(i32 42, i64 100, i32 %x), "foo"(i32 42, float 0.000000e+00, i32 %l) ]
; CHECK: call void @op_bundle_callee_1(i32 10, i32 %x) [ "foo"(i32 42, i64 100, i32 %x), "foo"(i32 42, float 0.000000e+00, i32 %l) ]
ret void
}
; Invoke versions of the above tests:
define void @invoke_with_operand_bundle0(i32* %ptr) personality i8 3 {
; CHECK-LABEL: @invoke_with_operand_bundle0(
entry:
%l = load i32, i32* %ptr
%x = add i32 42, 1
invoke void @op_bundle_callee_0() [ "foo"(i32 42, i64 100, i32 %x), "bar"(float 0.000000e+00, i64 100, i32 %l) ] to label %normal unwind label %exception
; CHECK: invoke void @op_bundle_callee_0() [ "foo"(i32 42, i64 100, i32 %x), "bar"(float 0.000000e+00, i64 100, i32 %l) ]
exception:
%cleanup = landingpad i8 cleanup
br label %normal
normal:
ret void
}
define void @invoke_with_operand_bundle1(i32* %ptr) personality i8 3 {
; CHECK-LABEL: @invoke_with_operand_bundle1(
entry:
%l = load i32, i32* %ptr
%x = add i32 42, 1
invoke void @op_bundle_callee_0() to label %normal unwind label %exception
; CHECK: invoke void @op_bundle_callee_0(){{$}}
exception:
%cleanup = landingpad i8 cleanup
br label %normal
normal:
invoke void @op_bundle_callee_0() [ "foo"() ] to label %normal1 unwind label %exception1
; CHECK: invoke void @op_bundle_callee_0() [ "foo"() ]
exception1:
%cleanup1 = landingpad i8 cleanup
br label %normal1
normal1:
invoke void @op_bundle_callee_0() [ "foo"(i32 42, i64 100, i32 %x), "foo"(i32 42, float 0.000000e+00, i32 %l) ] to label %normal2 unwind label %exception2
; CHECK: invoke void @op_bundle_callee_0() [ "foo"(i32 42, i64 100, i32 %x), "foo"(i32 42, float 0.000000e+00, i32 %l) ]
exception2:
%cleanup2 = landingpad i8 cleanup
br label %normal2
normal2:
ret void
}
define void @invoke_with_operand_bundle2(i32* %ptr) personality i8 3 {
; CHECK-LABEL: @invoke_with_operand_bundle2(
entry:
invoke void @op_bundle_callee_0() [ "foo"() ] to label %normal unwind label %exception
; CHECK: invoke void @op_bundle_callee_0() [ "foo"() ]
exception:
%cleanup = landingpad i8 cleanup
br label %normal
normal:
ret void
}
define void @invoke_with_operand_bundle3(i32* %ptr) personality i8 3 {
; CHECK-LABEL: @invoke_with_operand_bundle3(
entry:
%l = load i32, i32* %ptr
%x = add i32 42, 1
invoke void @op_bundle_callee_0() [ "foo"(i32 42, i64 100, i32 %x), "foo"(i32 42, float 0.000000e+00, i32 %l) ] to label %normal unwind label %exception
; CHECK: invoke void @op_bundle_callee_0() [ "foo"(i32 42, i64 100, i32 %x), "foo"(i32 42, float 0.000000e+00, i32 %l) ]
exception:
%cleanup = landingpad i8 cleanup
br label %normal
normal:
ret void
}
define void @invoke_with_operand_bundle4(i32* %ptr) personality i8 3 {
; CHECK-LABEL: @invoke_with_operand_bundle4(
entry:
%l = load i32, i32* %ptr
%x = add i32 42, 1
invoke void @op_bundle_callee_1(i32 10, i32 %x) [ "foo"(i32 42, i64 100, i32 %x), "foo"(i32 42, float 0.000000e+00, i32 %l) ]
to label %normal unwind label %exception
; CHECK: invoke void @op_bundle_callee_1(i32 10, i32 %x) [ "foo"(i32 42, i64 100, i32 %x), "foo"(i32 42, float 0.000000e+00, i32 %l) ]
exception:
%cleanup = landingpad i8 cleanup
br label %normal
normal:
ret void
}
BitcodeWriter: fix emission of invoke when calling a var-arg function with operand bundles llvm-svn: 281940
2016-09-20 05:27:04 +08:00
declare void @vaargs_func(...)
define void @invoke_with_operand_bundle_vaarg(i32* %ptr) personality i8 3 {
; CHECK-LABEL: @invoke_with_operand_bundle_vaarg(
entry:
%l = load i32, i32* %ptr
%x = add i32 42, 1
invoke void (...) @vaargs_func(i32 10, i32 %x) [ "foo"(i32 42, i64 100, i32 %x), "foo"(i32 42, float 0.000000e+00, i32 %l) ]
to label %normal unwind label %exception
; CHECK: invoke void (...) @vaargs_func(i32 10, i32 %x) [ "foo"(i32 42, i64 100, i32 %x), "foo"(i32 42, float 0.000000e+00, i32 %l) ]
exception:
%cleanup = landingpad i8 cleanup
br label %normal
normal:
ret void
}
Add writeonly IR attribute Summary: This complements the earlier addition of IntrWriteMem and IntrWriteArgMem LLVM intrinsic properties, see D18291. Also start using the attribute for memset, memcpy, and memmove intrinsics, and remove their special-casing in BasicAliasAnalysis. Reviewers: reames, joker.eph Subscribers: joker.eph, llvm-commits Differential Revision: http://reviews.llvm.org/D18714 llvm-svn: 274485
2016-07-04 16:01:29 +08:00
declare void @f.writeonly() writeonly
[FunctionAttrs] Annotate "willreturn" for intrinsics Summary: In D62801, new function attribute `willreturn` was introduced. In short, a function with `willreturn` is guaranteed to come back to the call site(more precise definition is in LangRef). In this patch, willreturn is annotated for LLVM intrinsics. Reviewers: jdoerfert Reviewed By: jdoerfert Subscribers: jvesely, nhaehnle, sstefan1, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D64904 llvm-svn: 367184
2019-07-28 14:09:56 +08:00
; CHECK: declare void @f.writeonly() #41
[Bitcode][Asm] Teach LLVM to read and write operand bundles. Summary: This also adds the first set of tests for operand bundles. The optimizer has not been audited to ensure that it does the right thing with operand bundles. Depends on D12456. Reviewers: reames, chandlerc, majnemer, dexonsmith, kmod, JosephTremoulet, rnk, bogner Subscribers: maksfb, llvm-commits Differential Revision: http://reviews.llvm.org/D12457 llvm-svn: 248551
2015-09-25 07:34:52 +08:00
Add speculatable function attribute This attribute tells the optimizer that the function may be speculated. Patch by Tom Stellard llvm-svn: 301680
2017-04-29 04:25:27 +08:00
declare void @f.speculatable() speculatable
[FunctionAttrs] Annotate "willreturn" for intrinsics Summary: In D62801, new function attribute `willreturn` was introduced. In short, a function with `willreturn` is guaranteed to come back to the call site(more precise definition is in LangRef). In this patch, willreturn is annotated for LLVM intrinsics. Reviewers: jdoerfert Reviewed By: jdoerfert Subscribers: jvesely, nhaehnle, sstefan1, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D64904 llvm-svn: 367184
2019-07-28 14:09:56 +08:00
; CHECK: declare void @f.speculatable() #42
Add speculatable function attribute This attribute tells the optimizer that the function may be speculated. Patch by Tom Stellard llvm-svn: 301680
2017-04-29 04:25:27 +08:00
IR: Introduce inrange attribute on getelementptr indices. If the inrange keyword is present before any index, loading from or storing to any pointer derived from the getelementptr has undefined behavior if the load or store would access memory outside of the bounds of the element selected by the index marked as inrange. This can be used, e.g. for alias analysis or to split globals at element boundaries where beneficial. As previously proposed on llvm-dev: http://lists.llvm.org/pipermail/llvm-dev/2016-July/102472.html Differential Revision: https://reviews.llvm.org/D22793 llvm-svn: 286514
2016-11-11 06:34:55 +08:00
;; Constant Expressions
define i8** @constexpr() {
; CHECK: ret i8** getelementptr inbounds ({ [4 x i8*], [4 x i8*] }, { [4 x i8*], [4 x i8*] }* null, i32 0, inrange i32 1, i32 2)
ret i8** getelementptr inbounds ({ [4 x i8*], [4 x i8*] }, { [4 x i8*], [4 x i8*] }* null, i32 0, inrange i32 1, i32 2)
}
Minor movement of one function with now-correct strictfp attribute to pass against the latest version of D68233.
2019-10-24 00:12:56 +08:00
define void @instructions.strictfp() strictfp {
call void @f.strictfp() strictfp
; CHECK: call void @f.strictfp() #43
ret void
}
[IR] Add Freeze instruction Summary: - Define Instruction::Freeze, let it be UnaryOperator - Add support for freeze to LLLexer/LLParser/BitcodeReader/BitcodeWriter The format is `%x = freeze <ty> %v` - Add support for freeze instruction to llvm-c interface. - Add m_Freeze in PatternMatch. - Erase freeze when lowering IR to SelDag. Reviewers: deadalnix, hfinkel, efriedma, lebedev.ri, nlopes, jdoerfert, regehr, filcab, delcypher, whitequark Reviewed By: lebedev.ri, jdoerfert Subscribers: jfb, kristof.beyls, hiraditya, lebedev.ri, steven_wu, dexonsmith, xbolva00, delcypher, spatel, regehr, trentxintong, vsk, filcab, nlopes, mehdi_amini, deadalnix, llvm-commits Differential Revision: https://reviews.llvm.org/D29011
2019-11-05 14:53:22 +08:00
define i64 @constexpr_freeze() {
ret i64 freeze (i64 32)
}
IR: Add immarg attribute This indicates an intrinsic parameter is required to be a constant, and should not be replaced with a non-constant value. Add the attribute to all AMDGPU and generic intrinsics that comments indicate it should apply to. I scanned other target intrinsics, but I don't see any obvious comments indicating which arguments are intended to be only immediates. This breaks one questionable testcase for the autoupgrade. I'm unclear on whether the autoupgrade is supposed to really handle declarations which were never valid. The verifier fails because the attributes now refer to a parameter past the end of the argument list. llvm-svn: 355981
2019-03-13 05:02:54 +08:00
; immarg attribute
declare void @llvm.test.immarg.intrinsic(i32 immarg)
; CHECK: declare void @llvm.test.immarg.intrinsic(i32 immarg)
Reapply: IR: add optional type to 'byval' function parameters When we switch to opaque pointer types we will need some way to describe how many bytes a 'byval' parameter should occupy on the stack. This adds a (for now) optional extra type parameter. If present, the type must match the pointee type of the argument. The original commit did not remap byval types when linking modules, which broke LTO. This version fixes that. Note to front-end maintainers: if this causes test failures, it's probably because the "byval" attribute is printed after attributes without any parameter after this change. llvm-svn: 362128
2019-05-31 02:48:23 +08:00
; byval attribute with type
%named_type = type [8 x i8]
declare void @byval_type(i32* byval(i32) align 2)
declare void @byval_type2({ i8, i8* }* byval({ i8, i8* }))
declare void @byval_named_type(%named_type* byval(%named_type))
; CHECK: declare void @byval_type(i32* byval(i32) align 2)
; CHECK: declare void @byval_type2({ i8, i8* }* byval({ i8, i8* }))
; CHECK: declare void @byval_named_type([8 x i8]* byval([8 x i8]))
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
; CHECK: attributes #0 = { alignstack=4 }
; CHECK: attributes #1 = { alignstack=8 }
; CHECK: attributes #2 = { alwaysinline }
; CHECK: attributes #3 = { cold }
; CHECK: attributes #4 = { convergent }
; CHECK: attributes #5 = { inlinehint }
; CHECK: attributes #6 = { jumptable }
; CHECK: attributes #7 = { minsize }
; CHECK: attributes #8 = { naked }
; CHECK: attributes #9 = { nobuiltin }
; CHECK: attributes #10 = { noduplicate }
; CHECK: attributes #11 = { noimplicitfloat }
; CHECK: attributes #12 = { noinline }
; CHECK: attributes #13 = { nonlazybind }
; CHECK: attributes #14 = { noredzone }
; CHECK: attributes #15 = { noreturn }
; CHECK: attributes #16 = { nounwind }
; CHECK: attributes #17 = { noinline optnone }
; CHECK: attributes #18 = { optsize }
; CHECK: attributes #19 = { readnone }
; CHECK: attributes #20 = { readonly }
; CHECK: attributes #21 = { returns_twice }
; CHECK: attributes #22 = { safestack }
; CHECK: attributes #23 = { sanitize_address }
; CHECK: attributes #24 = { sanitize_memory }
; CHECK: attributes #25 = { sanitize_thread }
; CHECK: attributes #26 = { ssp }
; CHECK: attributes #27 = { sspreq }
; CHECK: attributes #28 = { sspstrong }
; CHECK: attributes #29 = { "thunk" }
; CHECK: attributes #30 = { uwtable }
; CHECK: attributes #31 = { "cpu"="cortex-a8" }
Add a new attribute: norecurse This attribute allows the compiler to assume that the function never recurses into itself, either directly or indirectly (transitively). This can be used among other things to demote global variables to locals. llvm-svn: 252282
2015-11-06 18:32:53 +08:00
; CHECK: attributes #32 = { norecurse }
Add InaccessibleMemOnly and inaccessibleMemOrArgMemOnly attributes Summary: This patch introduces two new function attributes InaccessibleMemOnly: This attribute indicates that the function may only access memory that is not accessible by the program/IR being compiled. This is a weaker form of ReadNone. inaccessibleMemOrArgMemOnly: This attribute indicates that the function may only access memory that is either not accessible by the program/IR being compiled, or is pointed to by its pointer arguments. This is a weaker form of ArgMemOnly Test cases have been updated. This revision uses this (https://github.com/llvm-mirror/llvm/commit/d001932f3a8aa1ebd1555162fdce365f011bc292) as reference. Reviewers: jmolloy, hfinkel Subscribers: reames, joker.eph, llvm-commits Differential Revision: http://reviews.llvm.org/D15499 llvm-svn: 255778
2015-12-17 00:16:19 +08:00
; CHECK: attributes #33 = { inaccessiblememonly }
; CHECK: attributes #34 = { inaccessiblemem_or_argmemonly }
[FunctionAttrs] Annotate "willreturn" for intrinsics Summary: In D62801, new function attribute `willreturn` was introduced. In short, a function with `willreturn` is guaranteed to come back to the call site(more precise definition is in LangRef). In this patch, willreturn is annotated for LLVM intrinsics. Reviewers: jdoerfert Reviewed By: jdoerfert Subscribers: jvesely, nhaehnle, sstefan1, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D64904 llvm-svn: 367184
2019-07-28 14:09:56 +08:00
; CHECK: attributes #35 = { nounwind readnone willreturn }
Add InaccessibleMemOnly and inaccessibleMemOrArgMemOnly attributes Summary: This patch introduces two new function attributes InaccessibleMemOnly: This attribute indicates that the function may only access memory that is not accessible by the program/IR being compiled. This is a weaker form of ReadNone. inaccessibleMemOrArgMemOnly: This attribute indicates that the function may only access memory that is either not accessible by the program/IR being compiled, or is pointed to by its pointer arguments. This is a weaker form of ArgMemOnly Test cases have been updated. This revision uses this (https://github.com/llvm-mirror/llvm/commit/d001932f3a8aa1ebd1555162fdce365f011bc292) as reference. Reviewers: jmolloy, hfinkel Subscribers: reames, joker.eph, llvm-commits Differential Revision: http://reviews.llvm.org/D15499 llvm-svn: 255778
2015-12-17 00:16:19 +08:00
; CHECK: attributes #36 = { argmemonly nounwind readonly }
; CHECK: attributes #37 = { argmemonly nounwind }
[FunctionAttrs] Annotate "willreturn" for intrinsics Summary: In D62801, new function attribute `willreturn` was introduced. In short, a function with `willreturn` is guaranteed to come back to the call site(more precise definition is in LangRef). In this patch, willreturn is annotated for LLVM intrinsics. Reviewers: jdoerfert Reviewed By: jdoerfert Subscribers: jvesely, nhaehnle, sstefan1, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D64904 llvm-svn: 367184
2019-07-28 14:09:56 +08:00
; CHECK: attributes #38 = { nounwind readnone }
; CHECK: attributes #39 = { nounwind readonly }
; CHECK: attributes #40 = { inaccessiblemem_or_argmemonly nounwind willreturn }
; CHECK: attributes #41 = { writeonly }
; CHECK: attributes #42 = { speculatable }
Minor movement of one function with now-correct strictfp attribute to pass against the latest version of D68233.
2019-10-24 00:12:56 +08:00
; CHECK: attributes #43 = { strictfp }
; CHECK: attributes #44 = { builtin }
IR: Add a broad bitcode compatibility test Successive versions of LLVM should retain the ability to parse bitcode generated by old releases of the compiler. This adds a bitcode format compatibility test, which is intended to provide good (albeit not entirely exhaustive) coverage of the current LangRef. This also includes compatibility tests for LLVM 3.6. After every 3.X.0 release, the compatibility.ll file from the 3.X branch should be copied to compatibility-3.X.ll on trunk, and the 3.X.0 release used to generate a corresponding bitcode file. Patch by Vedant Kumar! llvm-svn: 243779
2015-08-01 04:44:32 +08:00
;; Metadata
; Metadata -- Module flags
!llvm.module.flags = !{!0, !1, !2, !4, !5, !6}
; CHECK: !llvm.module.flags = !{!0, !1, !2, !4, !5, !6}
!0 = !{i32 1, !"mod1", i32 0}
; CHECK: !0 = !{i32 1, !"mod1", i32 0}
!1 = !{i32 2, !"mod2", i32 0}
; CHECK: !1 = !{i32 2, !"mod2", i32 0}
!2 = !{i32 3, !"mod3", !3}
; CHECK: !2 = !{i32 3, !"mod3", !3}
!3 = !{!"mod6", !0}
; CHECK: !3 = !{!"mod6", !0}
!4 = !{i32 4, !"mod4", i32 0}
; CHECK: !4 = !{i32 4, !"mod4", i32 0}
!5 = !{i32 5, !"mod5", !0}
; CHECK: !5 = !{i32 5, !"mod5", !0}
!6 = !{i32 6, !"mod6", !0}
; CHECK: !6 = !{i32 6, !"mod6", !0}
; Metadata -- Check `distinct'
!11 = distinct !{}
; CHECK: !11 = distinct !{}
!12 = distinct !{}
; CHECK: !12 = distinct !{}
!13 = !{!11}
; CHECK: !13 = !{!11}
!14 = !{!12}
; CHECK: !14 = !{!12}