maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
261,278 Commits 23 Branches 0 Tags 2.2 GiB
Commit Graph

31 Commits

Author SHA1 Message Date
Geoff Berry 92a286ae5a [SelectionDAG] Handle inverted conditions when splitting into multiple branches. 
Summary:
When conditional branches with complex conditions are split into
multiple branches in SelectionDAGBuilder::FindMergedConditions, also
handle inverted conditions.  These may sometimes appear without having
been optimized by InstCombine when CodeGenPrepare decides to sink and
duplicate cmp instructions, causing them to have only one use.  This
problem can be increased by e.g. GVNHoist hiding more cmps from
InstCombine by combining equivalent cmps from different blocks.

For example codegen X & !(Y | Z) as:
    jmp_if_X TmpBB
    jmp FBB
  TmpBB:
    jmp_if_notY Tmp2BB
    jmp FBB
  Tmp2BB:
    jmp_if_notZ TBB
    jmp FBB

Reviewers: bogner, MatzeB, qcolombet

Subscribers: llvm-commits, hiraditya, mcrosier, sebpop

Differential Revision: https://reviews.llvm.org/D28380

llvm-svn: 292944
 2017-01-24 16:36:07 +00:00
Sanjay Patel bf51c8a975 [x86] fix usage of stale operands when lowering select 
I noticed this problem as part of the ongoing attempt to canonicalize min/max ops in IR.

The debug output shows nodes like this:

t4: i32 = xor t2, Constant:i32<-1>
    t21: i8 = setcc t4, Constant:i32<0>, setlt:ch
  t14: i32 = select t21, t4, Constant:i32<-1>

And because the select is holding onto the t4 (xor) node while EmitTest creates a new 
x86-specific xor node, the lowering results in:

  t4: i32 = xor t2, Constant:i32<-1>
  t25: i32,i32 = X86ISD::XOR t2, Constant:i32<-1>
t28: i32,glue = X86ISD::CMOV Constant:i32<-1>, t4, Constant:i8<15>, t25:1

Differential Revision: https://reviews.llvm.org/D28374

llvm-svn: 291392
 2017-01-08 15:53:40 +00:00
Sanjay Patel 686527c1e0 [x86] add test to show bug in select lowering; NFC 
llvm-svn: 291151
 2017-01-05 18:35:44 +00:00
David L Kreitzer 8b959e5cfa Avoid unnecessary 32-bit to 64-bit zero extensions following 
32-bit CMOV instructions on x86_64. The 32-bit CMOV implicitly
zero extends.

Differential Revision: https://reviews.llvm.org/D22941

llvm-svn: 277148
 2016-07-29 15:09:54 +00:00
Michael Kuperstein 3e3652aef2 Recommit r274692 - [X86] Transform setcc + movzbl into xorl + setcc 
xorl + setcc is generally the preferred sequence due to the partial register
stall setcc + movzbl suffers from. As a bonus, it also encodes one byte smaller.
This fixes PR28146.

The original commit tried inserting an 8bit-subreg into a GR32 (not GR32_ABCD)
which was not appreciated by fast regalloc on 32-bit.

llvm-svn: 274802
 2016-07-07 22:50:23 +00:00
Michael Kuperstein edb38a94f8 Revert r274692 to check whether this is what breaks windows selfhost. 
llvm-svn: 274771
 2016-07-07 16:55:35 +00:00
Michael Kuperstein 1ef6c59b1d [X86] Transform setcc + movzbl into xorl + setcc 
xorl + setcc is generally the preferred sequence due to the partial register
stall setcc + movzbl suffers from. As a bonus, it also encodes one byte smaller.

This fixes PR28146.

Differential Revision: http://reviews.llvm.org/D21774

llvm-svn: 274692
 2016-07-06 21:56:18 +00:00
David Blaikie 23af64846f [opaque pointer type] Add textual IR support for explicit type parameter to the call instruction 
See r230786 and r230794 for similar changes to gep and load
respectively.

Call is a bit different because it often doesn't have a single explicit
type - usually the type is deduced from the arguments, and just the
return type is explicit. In those cases there's no need to change the
IR.

When that's not the case, the IR usually contains the pointer type of
the first operand - but since typed pointers are going away, that
representation is insufficient so I'm just stripping the "pointerness"
of the explicit type away.

This does make the IR a bit weird - it /sort of/ reads like the type of
the first operand: "call void () %x(" but %x is actually of type "void
()*" and will eventually be just of type "ptr". But this seems not too
bad and I don't think it would benefit from repeating the type
("void (), void () * %x(" and then eventually "void (), ptr %x(") as has
been done with gep and load.

This also has a side benefit: since the explicit type is no longer a
pointer, there's no ambiguity between an explicit type and a function
that returns a function pointer. Previously this case needed an explicit
type (eg: a function returning a void() function was written as
"call void () () * @x(" rather than "call void () * @x(" because of the
ambiguity between a function returning a pointer to a void() function
and a function returning void).

No ambiguity means even function pointer return types can just be
written alone, without writing the whole function's type.

This leaves /only/ the varargs case where the explicit type is required.

Given the special type syntax in call instructions, the regex-fu used
for migration was a bit more involved in its own unique way (as every
one of these is) so here it is. Use it in conjunction with the apply.sh
script and associated find/xargs commands I've provided in rr230786 to
migrate your out of tree tests. Do let me know if any of this doesn't
cover your cases & we can iterate on a more general script/regexes to
help others with out of tree tests.

About 9 test cases couldn't be automatically migrated - half of those
were functions returning function pointers, where I just had to manually
delete the function argument types now that we didn't need an explicit
function type there. The other half were typedefs of function types used
in calls - just had to manually drop the * from those.

import fileinput
import sys
import re

pat = re.compile(r'((?:=|:|^|\s)call\s(?:[^@]*?))(\s*$|\s*(?:(?:\[\[[a-zA-Z0-9_]+\]\]|[@%](?:(")?[\\\?@a-zA-Z0-9_.]*?(?(3)"|)|{{.*}}))(?:\(|$)|undef|inttoptr|bitcast|null|asm).*$)')
addrspace_end = re.compile(r"addrspace\(\d+\)\s*\*$")
func_end = re.compile("(?:void.*|\)\s*)\*$")

def conv(match, line):
  if not match or re.search(addrspace_end, match.group(1)) or not re.search(func_end, match.group(1)):
    return line
  return line[:match.start()] + match.group(1)[:match.group(1).rfind('*')].rstrip() + match.group(2) + line[match.end():]

for line in sys.stdin:
  sys.stdout.write(conv(re.search(pat, line), line))

llvm-svn: 235145
 2015-04-16 23:24:18 +00:00
David Blaikie f72d05bc7b [opaque pointer type] Add textual IR support for explicit type parameter to gep operator 
Similar to gep (r230786) and load (r230794) changes.

Similar migration script can be used to update test cases, which
successfully migrated all of LLVM and Polly, but about 4 test cases
needed manually changes in Clang.

(this script will read the contents of stdin and massage it into stdout
- wrap it in the 'apply.sh' script shown in previous commits + xargs to
apply it over a large set of test cases)

import fileinput
import sys
import re

rep = re.compile(r"(getelementptr(?:\s+inbounds)?\s*\()((<\d*\s+x\s+)?([^@]*?)(|\s*addrspace\(\d+\))\s*\*(?(3)>)\s*)(?=$|%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|zeroinitializer|<|\[\[[a-zA-Z]|\{\{)", re.MULTILINE | re.DOTALL)

def conv(match):
  line = match.group(1)
  line += match.group(4)
  line += ", "
  line += match.group(2)
  return line

line = sys.stdin.read()
off = 0
for match in re.finditer(rep, line):
  sys.stdout.write(line[off:match.start()])
  sys.stdout.write(conv(match))
  off = match.end()
sys.stdout.write(line[off:])

llvm-svn: 232184
 2015-03-13 18:20:45 +00:00
David Blaikie a79ac14fa6 [opaque pointer type] Add textual IR support for explicit type parameter to load instruction 
Essentially the same as the GEP change in r230786.

A similar migration script can be used to update test cases, though a few more
test case improvements/changes were required this time around: (r229269-r229278)

import fileinput
import sys
import re

pat = re.compile(r"((?:=|:|^)\s*load (?:atomic )?(?:volatile )?(.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$)")

for line in sys.stdin:
  sys.stdout.write(re.sub(pat, r"\1, \2\3*\4", line))

Reviewers: rafael, dexonsmith, grosser

Differential Revision: http://reviews.llvm.org/D7649

llvm-svn: 230794
 2015-02-27 21:17:42 +00:00
JF Bastien 5d3280c7a7 x86-32: PUSHF/POPF use/def EFLAGS 
Summary: As a side-quest for D6629 jvoung pointed out that I should use -verify-machineinstrs and this found a bug in x86-32's handling of EFLAGS for PUSHF/POPF. This patch fixes the use/def, and adds -verify-machineinstrs to all x86 tests which contain 'EFLAGS'. One exception: this patch leaves inline-asm-fpstack.ll as-is because it fails -verify-machineinstrs in a way unrelated to EFLAGS. This patch also modifies cmpxchg-clobber-flags.ll along the lines of what D6629 already does by also testing i386.

Test Plan: ninja check

Reviewers: t.p.northover, jvoung

Subscribers: llvm-commits

Differential Revision: http://reviews.llvm.org/D6687

llvm-svn: 224359
 2014-12-16 20:15:45 +00:00
Andrew Trick e4083f9e85 Disabled subregister copy coalescing during MachineCSE. 
This effectively backs out r197465 but leaves some of the general
fixes in place. Not all targets are ready to handle this feature. To
enable it, some infrastructure work is needed to better handle
register class constraints.

llvm-svn: 197514
 2013-12-17 19:29:36 +00:00
Andrew Trick e339828b90 Allow MachineCSE to coalesce trivial subregister copies the same way that it coalesces normal copies. 
Without this, MachineCSE is powerless to handle redundant operations with truncated source operands.

This required fixing the 2-addr pass to handle tied subregisters. It isn't clear what combinations of subregisters can legally be tied, but the simple case of truncated source operands is now safely handled:

     %vreg11<def> = COPY %vreg1:sub_32bit; GR32:%vreg11 GR64:%vreg1
     %vreg12<def> = COPY %vreg2:sub_32bit; GR32:%vreg12 GR64:%vreg2
     %vreg13<def,tied1> = ADD32rr %vreg11<tied0>, %vreg12<kill>, %EFLAGS<imp-def>

Test case: cse-add-with-overflow.ll.

This exposed an existing bug in
PPCInstrInfo::commuteInstruction. Thanks to Rafael for the test case:
PowerPC/crash.ll.

llvm-svn: 197465
 2013-12-17 04:50:45 +00:00
Rafael Espindola f152836788 Revert "Allow MachineCSE to coalesce trivial subregister copies the same way that it coalesces normal copies." 
This reverts commit r197414.

It broke the ppc64 bootstrap. I will post a testcase in a sec.

llvm-svn: 197424
 2013-12-16 20:57:09 +00:00
Andrew Trick 88bd8629b2 Allow MachineCSE to coalesce trivial subregister copies the same way 
that it coalesces normal copies.

Without this, MachineCSE is powerless to handle redundant operations
with truncated source operands.

This required fixing the 2-addr pass to handle tied subregisters. It
isn't clear what combinations of subregisters can legally be tied, but
the simple case of truncated source operands is now safely handled:

     %vreg11<def> = COPY %vreg1:sub_32bit; GR32:%vreg11 GR64:%vreg1
     %vreg12<def> = COPY %vreg2:sub_32bit; GR32:%vreg12 GR64:%vreg2
     %vreg13<def,tied1> = ADD32rr %vreg11<tied0>, %vreg12<kill>, %EFLAGS<imp-def>

llvm-svn: 197414
 2013-12-16 19:36:21 +00:00
Andrew Trick e97d8d6dde Enable MI Sched for x86. 
This changes the SelectionDAG scheduling preference to source
order. Soon, the SelectionDAG scheduler can be bypassed saving
a nice chunk of compile time.

Performance differences that result from this change are often a
consequence of register coalescing. The register coalescer is far from
perfect. Bugs can be filed for deficiencies.

On x86 SandyBridge/Haswell, the source order schedule is often
preserved, particularly for small blocks.

Register pressure is generally improved over the SD scheduler's ILP
mode. However, we are still able to handle large blocks that require
latency hiding, unlike the SD scheduler's BURR mode. MI scheduler also
attempts to discover the critical path in single-block loops and
adjust heuristics accordingly.

The MI scheduler relies on the new machine model. This is currently
unimplemented for AVX, so we may not be generating the best code yet.

Unit tests are updated so they don't depend on SD scheduling heuristics.

llvm-svn: 192750
 2013-10-15 23:33:07 +00:00
Stephen Lin f799e3f944 Convert CodeGen/*/*.ll tests to use the new CHECK-LABEL for easier debugging. No functionality change and all tests pass after conversion. 
This was done with the following sed invocation to catch label lines demarking function boundaries:
    sed -i '' "s/^;\( *\)\([A-Z0-9_]*\):\( *\)test\([A-Za-z0-9_-]*\):\( *\)$/;\1\2-LABEL:\3test\4:\5/g" test/CodeGen/*/*.ll
which was written conservatively to avoid false positives rather than false negatives. I scanned through all the changes and everything looks correct.

llvm-svn: 186258
 2013-07-13 20:38:47 +00:00
Andrew Trick 121124acf8 Revert "Temporarily enable MI-Sched on X86." 
This reverts commit 98a9b72e8c56dc13a2617de84503a3d78352789c.

llvm-svn: 184823
 2013-06-25 02:48:58 +00:00
Andrew Trick 5a1e0af838 Temporarily enable MI-Sched on X86. 
Sorry for the unit test churn. I'll try to make the change permanently
next time.

llvm-svn: 184705
 2013-06-24 09:13:20 +00:00
Manman Ren cc1dc6dc11 Disable rematerialization in TwoAddressInstructionPass. 
It is redundant; RegisterCoalescer will do the remat if it can't eliminate
the copy. Collected instruction counts before and after this. A few extra
instructions are generated due to spilling but it is normal to see these kinds
of changes with almost any small codegen change, according to Jakob.

This also fixed rdar://11830760 where xor is expected instead of movi0.

llvm-svn: 160749
 2012-07-25 18:28:13 +00:00
Benjamin Kramer 3d38c17b59 Switch the select to branch transformation on by default. 
The primitive conservative heuristic seems to give a slight overall
improvement while not regressing stuff. Make it available to wider
testing. If you notice any speed regressions (or significant code
size regressions) let me know!

llvm-svn: 156258
 2012-05-06 14:25:16 +00:00
Chris Lattner 6a144a2227 Upgrade syntax of tests using volatile instructions to use 'load volatile' instead of 'volatile load', which is archaic. 
llvm-svn: 145171
 2011-11-27 06:54:59 +00:00
Jakob Stoklund Olesen 1f72dd40c7 Pseudo CMOV instructions don't clobber EFLAGS. 
The explanation about a 0 argument being materialized as xor is no
longer valid.  Rematerialization will check if EFLAGS is live before
clobbering it.

The code produced by X86TargetLowering::EmitLoweredSelect does not
clobber EFLAGS.

This causes one less testb instruction to be generated in the cmov.ll
test case.

llvm-svn: 139057
 2011-09-02 23:52:55 +00:00
Bill Wendling 410ec4aad1 As Dan pointed out, movzbl, movsbl, and friends are nicer than their alias 
(movzx/movsx) because they give more information. Revert that part of the patch.

llvm-svn: 129498
 2011-04-14 01:46:37 +00:00
Bill Wendling 7e07d6fb69 Have the X86 back-end emit the alias instead of what's being aliased. In most 
cases, it's much nicer and more informative reading the alias.

llvm-svn: 129497
 2011-04-14 01:11:51 +00:00
Sean Callanan 04d8cb74f3 Instruction fixes, added instructions, and AsmString changes in the 
X86 instruction tables.

Also (while I was at it) cleaned up the X86 tables, removing tabs and
80-line violations.

This patch was reviewed by Chris Lattner, but please let me know if
there are any problems.

* X86*.td
	Removed tabs and fixed 80-line violations

* X86Instr64bit.td
	(IRET, POPCNT, BT_, LSL, SWPGS, PUSH_S, POP_S, L_S, SMSW)
		Added
	(CALL, CMOV) Added qualifiers
	(JMP) Added PC-relative jump instruction
	(POPFQ/PUSHFQ) Added qualifiers; renamed PUSHFQ to indicate
		that it is 64-bit only (ambiguous since it has no
		REX prefix)
	(MOV) Added rr form going the other way, which is encoded
		differently
	(MOV) Changed immediates to offsets, which is more correct;
		also fixed MOV64o64a to have to a 64-bit offset
	(MOV) Fixed qualifiers
	(MOV) Added debug-register and condition-register moves
	(MOVZX) Added more forms
	(ADC, SUB, SBB, AND, OR, XOR) Added reverse forms, which
		(as with MOV) are encoded differently
	(ROL) Made REX.W required
	(BT) Uncommented mr form for disassembly only
	(CVT__2__) Added several missing non-intrinsic forms
	(LXADD, XCHG) Reordered operands to make more sense for
		MRMSrcMem
	(XCHG) Added register-to-register forms
	(XADD, CMPXCHG, XCHG) Added non-locked forms
* X86InstrSSE.td
	(CVTSS2SI, COMISS, CVTTPS2DQ, CVTPS2PD, CVTPD2PS, MOVQ)
		Added
* X86InstrFPStack.td
	(COM_FST0, COMP_FST0, COM_FI, COM_FIP, FFREE, FNCLEX, FNOP,
	 FXAM, FLDL2T, FLDL2E, FLDPI, FLDLG2, FLDLN2, F2XM1, FYL2X,
	 FPTAN, FPATAN, FXTRACT, FPREM1, FDECSTP, FINCSTP, FPREM,
	 FYL2XP1, FSINCOS, FRNDINT, FSCALE, FCOMPP, FXSAVE,
	 FXRSTOR)
		Added
	(FCOM, FCOMP) Added qualifiers
	(FSTENV, FSAVE, FSTSW) Fixed opcode names
	(FNSTSW) Added implicit register operand
* X86InstrInfo.td
	(opaque512mem) Added for FXSAVE/FXRSTOR
	(offset8, offset16, offset32, offset64) Added for MOV
	(NOOPW, IRET, POPCNT, IN, BTC, BTR, BTS, LSL, INVLPG, STR,
	 LTR, PUSHFS, PUSHGS, POPFS, POPGS, LDS, LSS, LES, LFS,
	 LGS, VERR, VERW, SGDT, SIDT, SLDT, LGDT, LIDT, LLDT,
	 LODSD, OUTSB, OUTSW, OUTSD, HLT, RSM, FNINIT, CLC, STC,
	 CLI, STI, CLD, STD, CMC, CLTS, XLAT, WRMSR, RDMSR, RDPMC,
	 SMSW, LMSW, CPUID, INVD, WBINVD, INVEPT, INVVPID, VMCALL,
	 VMCLEAR, VMLAUNCH, VMRESUME, VMPTRLD, VMPTRST, VMREAD,
	 VMWRITE, VMXOFF, VMXON) Added
	(NOOPL, POPF, POPFD, PUSHF, PUSHFD) Added qualifier
	(JO, JNO, JB, JAE, JE, JNE, JBE, JA, JS, JNS, JP, JNP, JL,
	 JGE, JLE, JG, JCXZ) Added 32-bit forms
	(MOV) Changed some immediate forms to offset forms
	(MOV) Added reversed reg-reg forms, which are encoded
		differently
	(MOV) Added debug-register and condition-register moves
	(CMOV) Added qualifiers
	(AND, OR, XOR, ADC, SUB, SBB) Added reverse forms, like MOV
	(BT) Uncommented memory-register forms for disassembler
	(MOVSX, MOVZX) Added forms
	(XCHG, LXADD) Made operand order make sense for MRMSrcMem
	(XCHG) Added register-register forms
	(XADD, CMPXCHG) Added unlocked forms
* X86InstrMMX.td
	(MMX_MOVD, MMV_MOVQ) Added forms
* X86InstrInfo.cpp: Changed PUSHFQ to PUSHFQ64 to reflect table
	change

* X86RegisterInfo.td: Added debug and condition register sets
* x86-64-pic-3.ll: Fixed testcase to reflect call qualifier
* peep-test-3.ll: Fixed testcase to reflect test qualifier
* cmov.ll: Fixed testcase to reflect cmov qualifier
* loop-blocks.ll: Fixed testcase to reflect call qualifier
* x86-64-pic-11.ll: Fixed testcase to reflect call qualifier
* 2009-11-04-SubregCoalescingBug.ll: Fixed testcase to reflect call
  qualifier
* x86-64-pic-2.ll: Fixed testcase to reflect call qualifier
* live-out-reg-info.ll: Fixed testcase to reflect test qualifier
* tail-opts.ll: Fixed testcase to reflect call qualifiers
* x86-64-pic-10.ll: Fixed testcase to reflect call qualifier
* bss-pagealigned.ll: Fixed testcase to reflect call qualifier
* x86-64-pic-1.ll: Fixed testcase to reflect call qualifier
* widen_load-1.ll: Fixed testcase to reflect call qualifier

llvm-svn: 91638
 2009-12-18 00:01:26 +00:00
Dan Gohman d29cafc083 Restore a comment that was lost in the merge. 
llvm-svn: 81857
 2009-09-15 15:09:54 +00:00
Chris Lattner 30fceaba05 this is failing on linux hosts, force a triple. 
llvm-svn: 81833
 2009-09-15 04:27:29 +00:00
Chris Lattner 170c116aa7 merge one more in. 
llvm-svn: 81824
 2009-09-15 02:27:23 +00:00
Chris Lattner 42aaa6e443 merge some more cmov tests into cmov.ll 
llvm-svn: 81823
 2009-09-15 02:25:21 +00:00
Chris Lattner baf78e5b52 merge two cmov tests into one. 
llvm-svn: 81822
 2009-09-15 02:22:47 +00:00