maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/clang/test/CodeGen/mips-varargs.c

177 lines
6.6 KiB
C
Raw Normal View History

[mips] Fix r218248's testcase to use -O1 instead of -O3. llvm-svn: 218298
2014-09-23 16:58:04 +08:00
// RUN: %clang_cc1 -triple mips-unknown-linux -o - -O1 -emit-llvm %s | FileCheck %s -check-prefix=ALL -check-prefix=O32
Allow EmitVAArg() to promote types and use this to fix some N32/N64 vararg issues for Mips. Summary: With this patch, passing a va_list to another function and reading 10 int's from it works correctly on a big-endian target. Based on a pair of patches by David Chisnall, one of which I've reworked for the current trunk. Reviewers: theraven, atanasyan Reviewed By: theraven, atanasyan Subscribers: cfe-commits Differential Revision: http://reviews.llvm.org/D6248 llvm-svn: 222339
2014-11-19 18:01:35 +08:00
// RUN: %clang_cc1 -triple mipsel-unknown-linux -o - -O1 -emit-llvm %s | FileCheck %s -check-prefix=ALL -check-prefix=O32
// RUN: %clang_cc1 -triple mips64-unknown-linux -o - -O1 -emit-llvm -target-abi n32 %s | FileCheck %s -check-prefix=ALL -check-prefix=N32 -check-prefix=NEW
// RUN: %clang_cc1 -triple mips64-unknown-linux -o - -O1 -emit-llvm -target-abi n32 %s | FileCheck %s -check-prefix=ALL -check-prefix=N32 -check-prefix=NEW
// RUN: %clang_cc1 -triple mips64-unknown-linux -o - -O1 -emit-llvm %s | FileCheck %s -check-prefix=ALL -check-prefix=N64 -check-prefix=NEW
// RUN: %clang_cc1 -triple mips64el-unknown-linux -o - -O1 -emit-llvm %s | FileCheck %s -check-prefix=ALL -check-prefix=N64 -check-prefix=NEW
[mips] Correct alignment of vectors passed in varargs for the O32 ABI. Summary: Vectors are normally 16-byte aligned, however the O32 ABI enforces a maximum alignment of 8-bytes since the base of the stack is 8-byte aligned. Previously, this was enforced on the caller side, but not on the callee side. This fixes the output of OpenCL's printf when given vectors. Reviewers: atanasyan Reviewed By: atanasyan Subscribers: llvm-commits, pekka.jaaskelainen Differential Revision: http://reviews.llvm.org/D5433 llvm-svn: 218248
2014-09-22 21:27:06 +08:00
#include <stdarg.h>
typedef int v4i32 __attribute__ ((__vector_size__ (16)));
Allow EmitVAArg() to promote types and use this to fix some N32/N64 vararg issues for Mips. Summary: With this patch, passing a va_list to another function and reading 10 int's from it works correctly on a big-endian target. Based on a pair of patches by David Chisnall, one of which I've reworked for the current trunk. Reviewers: theraven, atanasyan Reviewed By: theraven, atanasyan Subscribers: cfe-commits Differential Revision: http://reviews.llvm.org/D6248 llvm-svn: 222339
2014-11-19 18:01:35 +08:00
int test_i32(char *fmt, ...) {
[mips] Correct alignment of vectors passed in varargs for the O32 ABI. Summary: Vectors are normally 16-byte aligned, however the O32 ABI enforces a maximum alignment of 8-bytes since the base of the stack is 8-byte aligned. Previously, this was enforced on the caller side, but not on the callee side. This fixes the output of OpenCL's printf when given vectors. Reviewers: atanasyan Reviewed By: atanasyan Subscribers: llvm-commits, pekka.jaaskelainen Differential Revision: http://reviews.llvm.org/D5433 llvm-svn: 218248
2014-09-22 21:27:06 +08:00
va_list va;
va_start(va, fmt);
Allow EmitVAArg() to promote types and use this to fix some N32/N64 vararg issues for Mips. Summary: With this patch, passing a va_list to another function and reading 10 int's from it works correctly on a big-endian target. Based on a pair of patches by David Chisnall, one of which I've reworked for the current trunk. Reviewers: theraven, atanasyan Reviewed By: theraven, atanasyan Subscribers: cfe-commits Differential Revision: http://reviews.llvm.org/D6248 llvm-svn: 222339
2014-11-19 18:01:35 +08:00
int v = va_arg(va, int);
[mips] Correct alignment of vectors passed in varargs for the O32 ABI. Summary: Vectors are normally 16-byte aligned, however the O32 ABI enforces a maximum alignment of 8-bytes since the base of the stack is 8-byte aligned. Previously, this was enforced on the caller side, but not on the callee side. This fixes the output of OpenCL's printf when given vectors. Reviewers: atanasyan Reviewed By: atanasyan Subscribers: llvm-commits, pekka.jaaskelainen Differential Revision: http://reviews.llvm.org/D5433 llvm-svn: 218248
2014-09-22 21:27:06 +08:00
va_end(va);
Allow EmitVAArg() to promote types and use this to fix some N32/N64 vararg issues for Mips. Summary: With this patch, passing a va_list to another function and reading 10 int's from it works correctly on a big-endian target. Based on a pair of patches by David Chisnall, one of which I've reworked for the current trunk. Reviewers: theraven, atanasyan Reviewed By: theraven, atanasyan Subscribers: cfe-commits Differential Revision: http://reviews.llvm.org/D6248 llvm-svn: 222339
2014-11-19 18:01:35 +08:00
return v;
[mips] Correct alignment of vectors passed in varargs for the O32 ABI. Summary: Vectors are normally 16-byte aligned, however the O32 ABI enforces a maximum alignment of 8-bytes since the base of the stack is 8-byte aligned. Previously, this was enforced on the caller side, but not on the callee side. This fixes the output of OpenCL's printf when given vectors. Reviewers: atanasyan Reviewed By: atanasyan Subscribers: llvm-commits, pekka.jaaskelainen Differential Revision: http://reviews.llvm.org/D5433 llvm-svn: 218248
2014-09-22 21:27:06 +08:00
}
Allow EmitVAArg() to promote types and use this to fix some N32/N64 vararg issues for Mips. Summary: With this patch, passing a va_list to another function and reading 10 int's from it works correctly on a big-endian target. Based on a pair of patches by David Chisnall, one of which I've reworked for the current trunk. Reviewers: theraven, atanasyan Reviewed By: theraven, atanasyan Subscribers: cfe-commits Differential Revision: http://reviews.llvm.org/D6248 llvm-svn: 222339
2014-11-19 18:01:35 +08:00
// ALL-LABEL: define i32 @test_i32(i8*{{.*}} %fmt, ...)
[mips] Correct alignment of vectors passed in varargs for the O32 ABI. Summary: Vectors are normally 16-byte aligned, however the O32 ABI enforces a maximum alignment of 8-bytes since the base of the stack is 8-byte aligned. Previously, this was enforced on the caller side, but not on the callee side. This fixes the output of OpenCL's printf when given vectors. Reviewers: atanasyan Reviewed By: atanasyan Subscribers: llvm-commits, pekka.jaaskelainen Differential Revision: http://reviews.llvm.org/D5433 llvm-svn: 218248
2014-09-22 21:27:06 +08:00
//
// O32: %va = alloca i8*, align [[PTRALIGN:4]]
// N32: %va = alloca i8*, align [[PTRALIGN:4]]
// N64: %va = alloca i8*, align [[PTRALIGN:8]]
//
clang/test/CodeGen/mips-varargs.c: Fixup for -Asserts. llvm-svn: 218256
2014-09-23 00:40:05 +08:00
// ALL: [[VA1:%.+]] = bitcast i8** %va to i8*
// ALL: call void @llvm.va_start(i8* [[VA1]])
Allow EmitVAArg() to promote types and use this to fix some N32/N64 vararg issues for Mips. Summary: With this patch, passing a va_list to another function and reading 10 int's from it works correctly on a big-endian target. Based on a pair of patches by David Chisnall, one of which I've reworked for the current trunk. Reviewers: theraven, atanasyan Reviewed By: theraven, atanasyan Subscribers: cfe-commits Differential Revision: http://reviews.llvm.org/D6248 llvm-svn: 222339
2014-11-19 18:01:35 +08:00
//
// O32: [[TMP0:%.+]] = bitcast i8** %va to i32**
// O32: [[AP_CUR:%.+]] = load i32** [[TMP0]], align [[PTRALIGN]]
// NEW: [[TMP0:%.+]] = bitcast i8** %va to i64**
// NEW: [[AP_CUR:%.+]] = load i64** [[TMP0]], align [[PTRALIGN]]
//
Update Clang tests that run the LLVM optimizer to reflect the changed canonicalization in r222748. No interesting functionality changed here. llvm-svn: 222749
2014-11-25 18:10:37 +08:00
// O32: [[AP_NEXT:%.+]] = getelementptr i32* [[AP_CUR]], i32 1
// NEW: [[AP_NEXT:%.+]] = getelementptr i64* [[AP_CUR]], {{i32|i64}} 1
Allow EmitVAArg() to promote types and use this to fix some N32/N64 vararg issues for Mips. Summary: With this patch, passing a va_list to another function and reading 10 int's from it works correctly on a big-endian target. Based on a pair of patches by David Chisnall, one of which I've reworked for the current trunk. Reviewers: theraven, atanasyan Reviewed By: theraven, atanasyan Subscribers: cfe-commits Differential Revision: http://reviews.llvm.org/D6248 llvm-svn: 222339
2014-11-19 18:01:35 +08:00
//
Update Clang tests that run the LLVM optimizer to reflect the changed canonicalization in r222748. No interesting functionality changed here. llvm-svn: 222749
2014-11-25 18:10:37 +08:00
// O32: store i32* [[AP_NEXT]], i32** [[TMP0]], align [[PTRALIGN]]
// NEW: store i64* [[AP_NEXT]], i64** [[TMP0]], align [[PTRALIGN]]
Allow EmitVAArg() to promote types and use this to fix some N32/N64 vararg issues for Mips. Summary: With this patch, passing a va_list to another function and reading 10 int's from it works correctly on a big-endian target. Based on a pair of patches by David Chisnall, one of which I've reworked for the current trunk. Reviewers: theraven, atanasyan Reviewed By: theraven, atanasyan Subscribers: cfe-commits Differential Revision: http://reviews.llvm.org/D6248 llvm-svn: 222339
2014-11-19 18:01:35 +08:00
//
// O32: [[ARG1:%.+]] = load i32* [[AP_CUR]], align 4
// NEW: [[TMP2:%.+]] = load i64* [[AP_CUR]], align 8
// NEW: [[ARG1:%.+]] = trunc i64 [[TMP2]] to i32
//
// ALL: call void @llvm.va_end(i8* [[VA1]])
// ALL: ret i32 [[ARG1]]
// ALL: }
int test_i32_2args(char *fmt, ...) {
va_list va;
va_start(va, fmt);
int v1 = va_arg(va, int);
int v2 = va_arg(va, int);
va_end(va);
return v1 + v2;
}
// ALL-LABEL: define i32 @test_i32_2args(i8*{{.*}} %fmt, ...)
//
// ALL: %va = alloca i8*, align [[PTRALIGN]]
// ALL: [[VA1:%.+]] = bitcast i8** %va to i8*
// ALL: call void @llvm.va_start(i8* [[VA1]])
//
// O32: [[TMP0:%.+]] = bitcast i8** %va to i32**
// O32: [[AP_CUR:%.+]] = load i32** [[TMP0]], align [[PTRALIGN]]
// NEW: [[TMP0:%.+]] = bitcast i8** %va to i64**
// NEW: [[AP_CUR:%.+]] = load i64** [[TMP0]], align [[PTRALIGN]]
//
Update Clang tests that run the LLVM optimizer to reflect the changed canonicalization in r222748. No interesting functionality changed here. llvm-svn: 222749
2014-11-25 18:10:37 +08:00
// O32: [[AP_NEXT1:%.+]] = getelementptr i32* [[AP_CUR]], i32 1
// NEW: [[AP_NEXT1:%.+]] = getelementptr i64* [[AP_CUR]], [[INTPTR_T:i32|i64]] 1
Allow EmitVAArg() to promote types and use this to fix some N32/N64 vararg issues for Mips. Summary: With this patch, passing a va_list to another function and reading 10 int's from it works correctly on a big-endian target. Based on a pair of patches by David Chisnall, one of which I've reworked for the current trunk. Reviewers: theraven, atanasyan Reviewed By: theraven, atanasyan Subscribers: cfe-commits Differential Revision: http://reviews.llvm.org/D6248 llvm-svn: 222339
2014-11-19 18:01:35 +08:00
//
Update Clang tests that run the LLVM optimizer to reflect the changed canonicalization in r222748. No interesting functionality changed here. llvm-svn: 222749
2014-11-25 18:10:37 +08:00
// O32: store i32* [[AP_NEXT1]], i32** [[TMP0]], align [[PTRALIGN]]
Allow EmitVAArg() to promote types and use this to fix some N32/N64 vararg issues for Mips. Summary: With this patch, passing a va_list to another function and reading 10 int's from it works correctly on a big-endian target. Based on a pair of patches by David Chisnall, one of which I've reworked for the current trunk. Reviewers: theraven, atanasyan Reviewed By: theraven, atanasyan Subscribers: cfe-commits Differential Revision: http://reviews.llvm.org/D6248 llvm-svn: 222339
2014-11-19 18:01:35 +08:00
// FIXME: N32 optimised this store out. Why only for this ABI?
Update Clang tests that run the LLVM optimizer to reflect the changed canonicalization in r222748. No interesting functionality changed here. llvm-svn: 222749
2014-11-25 18:10:37 +08:00
// N64: store i64* [[AP_NEXT1]], i64** [[TMP0]], align [[PTRALIGN]]
Allow EmitVAArg() to promote types and use this to fix some N32/N64 vararg issues for Mips. Summary: With this patch, passing a va_list to another function and reading 10 int's from it works correctly on a big-endian target. Based on a pair of patches by David Chisnall, one of which I've reworked for the current trunk. Reviewers: theraven, atanasyan Reviewed By: theraven, atanasyan Subscribers: cfe-commits Differential Revision: http://reviews.llvm.org/D6248 llvm-svn: 222339
2014-11-19 18:01:35 +08:00
//
// O32: [[ARG1:%.+]] = load i32* [[AP_CUR]], align 4
// NEW: [[TMP3:%.+]] = load i64* [[AP_CUR]], align 8
// NEW: [[ARG1:%.+]] = trunc i64 [[TMP3]] to i32
//
Update Clang tests that run the LLVM optimizer to reflect the changed canonicalization in r222748. No interesting functionality changed here. llvm-svn: 222749
2014-11-25 18:10:37 +08:00
// O32: [[AP_NEXT2:%.+]] = getelementptr i32* [[AP_CUR]], i32 2
// NEW: [[AP_NEXT2:%.+]] = getelementptr i64* [[AP_CUR]], [[INTPTR_T]] 2
Allow EmitVAArg() to promote types and use this to fix some N32/N64 vararg issues for Mips. Summary: With this patch, passing a va_list to another function and reading 10 int's from it works correctly on a big-endian target. Based on a pair of patches by David Chisnall, one of which I've reworked for the current trunk. Reviewers: theraven, atanasyan Reviewed By: theraven, atanasyan Subscribers: cfe-commits Differential Revision: http://reviews.llvm.org/D6248 llvm-svn: 222339
2014-11-19 18:01:35 +08:00
//
Update Clang tests that run the LLVM optimizer to reflect the changed canonicalization in r222748. No interesting functionality changed here. llvm-svn: 222749
2014-11-25 18:10:37 +08:00
// O32: store i32* [[AP_NEXT2]], i32** [[TMP0]], align [[PTRALIGN]]
// NEW: store i64* [[AP_NEXT2]], i64** [[TMP0]], align [[PTRALIGN]]
Allow EmitVAArg() to promote types and use this to fix some N32/N64 vararg issues for Mips. Summary: With this patch, passing a va_list to another function and reading 10 int's from it works correctly on a big-endian target. Based on a pair of patches by David Chisnall, one of which I've reworked for the current trunk. Reviewers: theraven, atanasyan Reviewed By: theraven, atanasyan Subscribers: cfe-commits Differential Revision: http://reviews.llvm.org/D6248 llvm-svn: 222339
2014-11-19 18:01:35 +08:00
//
Update Clang tests that run the LLVM optimizer to reflect the changed canonicalization in r222748. No interesting functionality changed here. llvm-svn: 222749
2014-11-25 18:10:37 +08:00
// O32: [[ARG2:%.+]] = load i32* [[AP_NEXT1]], align 4
// NEW: [[TMP4:%.+]] = load i64* [[AP_NEXT1]], align 8
Allow EmitVAArg() to promote types and use this to fix some N32/N64 vararg issues for Mips. Summary: With this patch, passing a va_list to another function and reading 10 int's from it works correctly on a big-endian target. Based on a pair of patches by David Chisnall, one of which I've reworked for the current trunk. Reviewers: theraven, atanasyan Reviewed By: theraven, atanasyan Subscribers: cfe-commits Differential Revision: http://reviews.llvm.org/D6248 llvm-svn: 222339
2014-11-19 18:01:35 +08:00
// NEW: [[ARG2:%.+]] = trunc i64 [[TMP4]] to i32
//
// ALL: call void @llvm.va_end(i8* [[VA1]])
Revert "[Reassociate] Update test cases due to r222142." This reverts commit r222144. Commit r222142 is being reverted due to a spec2006/gcc execution-time regression. Update mips-varargs test as well. llvm-svn: 222397
2014-11-20 07:20:35 +08:00
// ALL: [[ADD:%.+]] = add nsw i32 [[ARG2]], [[ARG1]]
Allow EmitVAArg() to promote types and use this to fix some N32/N64 vararg issues for Mips. Summary: With this patch, passing a va_list to another function and reading 10 int's from it works correctly on a big-endian target. Based on a pair of patches by David Chisnall, one of which I've reworked for the current trunk. Reviewers: theraven, atanasyan Reviewed By: theraven, atanasyan Subscribers: cfe-commits Differential Revision: http://reviews.llvm.org/D6248 llvm-svn: 222339
2014-11-19 18:01:35 +08:00
// ALL: ret i32 [[ADD]]
// ALL: }
long long test_i64(char *fmt, ...) {
va_list va;
va_start(va, fmt);
long long v = va_arg(va, long long);
va_end(va);
return v;
}
// ALL-LABEL: define i64 @test_i64(i8*{{.*}} %fmt, ...)
//
// ALL: %va = alloca i8*, align [[PTRALIGN]]
// ALL: [[VA1:%.+]] = bitcast i8** %va to i8*
// ALL: call void @llvm.va_start(i8* [[VA1]])
//
// O32: [[AP_CUR:%.+]] = load i8** %va, align [[PTRALIGN]]
// NEW: [[TMP0:%.+]] = bitcast i8** %va to i64**
// NEW: [[AP_CUR:%.+]] = load i64** [[TMP0]], align [[PTRALIGN]]
//
// i64 is 8-byte aligned, while this is within O32's stack alignment there's no
// guarantee that the offset is still 8-byte aligned after earlier reads.
// O32: [[PTR0:%.+]] = ptrtoint i8* [[AP_CUR]] to [[INTPTR_T:i32]]
// O32: [[PTR1:%.+]] = add i32 [[PTR0]], 7
// O32: [[PTR2:%.+]] = and i32 [[PTR1]], -8
// O32: [[PTR3:%.+]] = inttoptr [[INTPTR_T]] [[PTR2]] to i64*
// O32: [[PTR4:%.+]] = inttoptr [[INTPTR_T]] [[PTR2]] to i8*
//
// O32: [[AP_NEXT:%.+]] = getelementptr i8* [[PTR4]], [[INTPTR_T]] 8
Update Clang tests that run the LLVM optimizer to reflect the changed canonicalization in r222748. No interesting functionality changed here. llvm-svn: 222749
2014-11-25 18:10:37 +08:00
// NEW: [[AP_NEXT:%.+]] = getelementptr i64* [[AP_CUR]], [[INTPTR_T:i32|i64]] 1
Allow EmitVAArg() to promote types and use this to fix some N32/N64 vararg issues for Mips. Summary: With this patch, passing a va_list to another function and reading 10 int's from it works correctly on a big-endian target. Based on a pair of patches by David Chisnall, one of which I've reworked for the current trunk. Reviewers: theraven, atanasyan Reviewed By: theraven, atanasyan Subscribers: cfe-commits Differential Revision: http://reviews.llvm.org/D6248 llvm-svn: 222339
2014-11-19 18:01:35 +08:00
//
Update Clang tests that run the LLVM optimizer to reflect the changed canonicalization in r222748. No interesting functionality changed here. llvm-svn: 222749
2014-11-25 18:10:37 +08:00
// O32: store i8* [[AP_NEXT]], i8** %va, align [[PTRALIGN]]
// NEW: store i64* [[AP_NEXT]], i64** [[TMP0]], align [[PTRALIGN]]
Allow EmitVAArg() to promote types and use this to fix some N32/N64 vararg issues for Mips. Summary: With this patch, passing a va_list to another function and reading 10 int's from it works correctly on a big-endian target. Based on a pair of patches by David Chisnall, one of which I've reworked for the current trunk. Reviewers: theraven, atanasyan Reviewed By: theraven, atanasyan Subscribers: cfe-commits Differential Revision: http://reviews.llvm.org/D6248 llvm-svn: 222339
2014-11-19 18:01:35 +08:00
//
// O32: [[ARG1:%.+]] = load i64* [[PTR3]], align 8
// NEW: [[ARG1:%.+]] = load i64* [[AP_CUR]], align 8
//
// ALL: call void @llvm.va_end(i8* [[VA1]])
// ALL: ret i64 [[ARG1]]
// ALL: }
int test_v4i32(char *fmt, ...) {
va_list va;
va_start(va, fmt);
v4i32 v = va_arg(va, v4i32);
va_end(va);
return v[0];
}
// ALL-LABEL: define i32 @test_v4i32(i8*{{.*}} %fmt, ...)
//
// ALL: %va = alloca i8*, align [[PTRALIGN]]
// ALL: [[VA1:%.+]] = bitcast i8** %va to i8*
// ALL: call void @llvm.va_start(i8* [[VA1]])
clang/test/CodeGen/mips-varargs.c: Fixup for -Asserts. llvm-svn: 218256
2014-09-23 00:40:05 +08:00
// ALL: [[AP_CUR:%.+]] = load i8** %va, align [[PTRALIGN]]
[mips] Correct alignment of vectors passed in varargs for the O32 ABI. Summary: Vectors are normally 16-byte aligned, however the O32 ABI enforces a maximum alignment of 8-bytes since the base of the stack is 8-byte aligned. Previously, this was enforced on the caller side, but not on the callee side. This fixes the output of OpenCL's printf when given vectors. Reviewers: atanasyan Reviewed By: atanasyan Subscribers: llvm-commits, pekka.jaaskelainen Differential Revision: http://reviews.llvm.org/D5433 llvm-svn: 218248
2014-09-22 21:27:06 +08:00
//
Allow EmitVAArg() to promote types and use this to fix some N32/N64 vararg issues for Mips. Summary: With this patch, passing a va_list to another function and reading 10 int's from it works correctly on a big-endian target. Based on a pair of patches by David Chisnall, one of which I've reworked for the current trunk. Reviewers: theraven, atanasyan Reviewed By: theraven, atanasyan Subscribers: cfe-commits Differential Revision: http://reviews.llvm.org/D6248 llvm-svn: 222339
2014-11-19 18:01:35 +08:00
// O32: [[PTR0:%.+]] = ptrtoint i8* [[AP_CUR]] to [[INTPTR_T:i32]]
// N32: [[PTR0:%.+]] = ptrtoint i8* [[AP_CUR]] to [[INTPTR_T:i32]]
// N64: [[PTR0:%.+]] = ptrtoint i8* [[AP_CUR]] to [[INTPTR_T:i64]]
[mips] Correct alignment of vectors passed in varargs for the O32 ABI. Summary: Vectors are normally 16-byte aligned, however the O32 ABI enforces a maximum alignment of 8-bytes since the base of the stack is 8-byte aligned. Previously, this was enforced on the caller side, but not on the callee side. This fixes the output of OpenCL's printf when given vectors. Reviewers: atanasyan Reviewed By: atanasyan Subscribers: llvm-commits, pekka.jaaskelainen Differential Revision: http://reviews.llvm.org/D5433 llvm-svn: 218248
2014-09-22 21:27:06 +08:00
//
// Vectors are 16-byte aligned, however the O32 ABI has a maximum alignment of
// 8-bytes since the base of the stack is 8-byte aligned.
clang/test/CodeGen/mips-varargs.c: Fixup for -Asserts. llvm-svn: 218256
2014-09-23 00:40:05 +08:00
// O32: [[PTR1:%.+]] = add i32 [[PTR0]], 7
// O32: [[PTR2:%.+]] = and i32 [[PTR1]], -8
[mips] Correct alignment of vectors passed in varargs for the O32 ABI. Summary: Vectors are normally 16-byte aligned, however the O32 ABI enforces a maximum alignment of 8-bytes since the base of the stack is 8-byte aligned. Previously, this was enforced on the caller side, but not on the callee side. This fixes the output of OpenCL's printf when given vectors. Reviewers: atanasyan Reviewed By: atanasyan Subscribers: llvm-commits, pekka.jaaskelainen Differential Revision: http://reviews.llvm.org/D5433 llvm-svn: 218248
2014-09-22 21:27:06 +08:00
//
Allow EmitVAArg() to promote types and use this to fix some N32/N64 vararg issues for Mips. Summary: With this patch, passing a va_list to another function and reading 10 int's from it works correctly on a big-endian target. Based on a pair of patches by David Chisnall, one of which I've reworked for the current trunk. Reviewers: theraven, atanasyan Reviewed By: theraven, atanasyan Subscribers: cfe-commits Differential Revision: http://reviews.llvm.org/D6248 llvm-svn: 222339
2014-11-19 18:01:35 +08:00
// NEW: [[PTR1:%.+]] = add [[INTPTR_T]] [[PTR0]], 15
// NEW: [[PTR2:%.+]] = and [[INTPTR_T]] [[PTR1]], -16
[mips] Correct alignment of vectors passed in varargs for the O32 ABI. Summary: Vectors are normally 16-byte aligned, however the O32 ABI enforces a maximum alignment of 8-bytes since the base of the stack is 8-byte aligned. Previously, this was enforced on the caller side, but not on the callee side. This fixes the output of OpenCL's printf when given vectors. Reviewers: atanasyan Reviewed By: atanasyan Subscribers: llvm-commits, pekka.jaaskelainen Differential Revision: http://reviews.llvm.org/D5433 llvm-svn: 218248
2014-09-22 21:27:06 +08:00
//
Allow EmitVAArg() to promote types and use this to fix some N32/N64 vararg issues for Mips. Summary: With this patch, passing a va_list to another function and reading 10 int's from it works correctly on a big-endian target. Based on a pair of patches by David Chisnall, one of which I've reworked for the current trunk. Reviewers: theraven, atanasyan Reviewed By: theraven, atanasyan Subscribers: cfe-commits Differential Revision: http://reviews.llvm.org/D6248 llvm-svn: 222339
2014-11-19 18:01:35 +08:00
// ALL: [[PTR3:%.+]] = inttoptr [[INTPTR_T]] [[PTR2]] to <4 x i32>*
// ALL: [[PTR4:%.+]] = inttoptr [[INTPTR_T]] [[PTR2]] to i8*
// ALL: [[AP_NEXT:%.+]] = getelementptr i8* [[PTR4]], [[INTPTR_T]] 16
clang/test/CodeGen/mips-varargs.c: Fixup for -Asserts. llvm-svn: 218256
2014-09-23 00:40:05 +08:00
// ALL: store i8* [[AP_NEXT]], i8** %va, align [[PTRALIGN]]
// ALL: [[PTR5:%.+]] = load <4 x i32>* [[PTR3]], align 16
// ALL: call void @llvm.va_end(i8* [[VA1]])
// ALL: [[VECEXT:%.+]] = extractelement <4 x i32> [[PTR5]], i32 0
// ALL: ret i32 [[VECEXT]]
[mips] Correct alignment of vectors passed in varargs for the O32 ABI. Summary: Vectors are normally 16-byte aligned, however the O32 ABI enforces a maximum alignment of 8-bytes since the base of the stack is 8-byte aligned. Previously, this was enforced on the caller side, but not on the callee side. This fixes the output of OpenCL's printf when given vectors. Reviewers: atanasyan Reviewed By: atanasyan Subscribers: llvm-commits, pekka.jaaskelainen Differential Revision: http://reviews.llvm.org/D5433 llvm-svn: 218248
2014-09-22 21:27:06 +08:00
// ALL: }