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

187 lines
5.9 KiB
LLVM
Raw Normal View History

Unpoison stack before resume instruction Summary: Clang inserts cleanup code before resume similar way as before return instruction. This makes asan poison local variables causing false use-after-scope reports. __asan_handle_no_return does not help here as it was executed before llvm.lifetime.end inserted into resume block. To avoid false report we need to unpoison stack for resume same way as for return. PR27453 Reviewers: kcc, eugenis Differential Revision: https://reviews.llvm.org/D22661 llvm-svn: 276480
2016-07-23 06:04:38 +08:00
; Test handling of llvm.lifetime intrinsics.
Revert "[asan] Avoid lifetime analysis for allocas with can be in ambiguous state" This approach is not good enough. Working on the new solution. This reverts commit r280907. llvm-svn: 281689
2016-09-16 09:38:46 +08:00
; RUN: opt < %s -asan -asan-module -asan-use-after-scope -asan-use-after-return=0 -S | FileCheck %s --check-prefixes=CHECK,CHECK-DEFAULT
; RUN: opt < %s -asan -asan-module -asan-use-after-scope -asan-use-after-return=0 -asan-instrument-dynamic-allocas=0 -S | FileCheck %s --check-prefixes=CHECK,CHECK-NO-DYNAMIC
ASan: add initial support for handling llvm.lifetime intrinsics in ASan - emit calls into runtime library that poison memory for local variables when their lifetime is over and unpoison memory when their lifetime begins. llvm-svn: 169200
2012-12-04 09:34:23 +08:00
target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64-S128"
target triple = "x86_64-unknown-linux-gnu"
Add address space mangling to lifetime intrinsics In preparation for allowing allocas to have non-0 addrspace. llvm-svn: 299876
2017-04-11 04:18:21 +08:00
declare void @llvm.lifetime.start.p0i8(i64, i8* nocapture) nounwind
declare void @llvm.lifetime.end.p0i8(i64, i8* nocapture) nounwind
ASan: add initial support for handling llvm.lifetime intrinsics in ASan - emit calls into runtime library that poison memory for local variables when their lifetime is over and unpoison memory when their lifetime begins. llvm-svn: 169200
2012-12-04 09:34:23 +08:00
Unify clang/llvm attributes for asan/tsan/msan (LLVM part) These are two related changes (one in llvm, one in clang). LLVM: - rename address_safety => sanitize_address (the enum value is the same, so we preserve binary compatibility with old bitcode) - rename thread_safety => sanitize_thread - rename no_uninitialized_checks -> sanitize_memory CLANG: - add __attribute__((no_sanitize_address)) as a synonym for __attribute__((no_address_safety_analysis)) - add __attribute__((no_sanitize_thread)) - add __attribute__((no_sanitize_memory)) for S in address thread memory If -fsanitize=S is present and __attribute__((no_sanitize_S)) is not set llvm attribute sanitize_S llvm-svn: 176075
2013-02-26 14:58:09 +08:00
define void @lifetime_no_size() sanitize_address {
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; CHECK-LABEL: define void @lifetime_no_size()
ASan: add initial support for handling llvm.lifetime intrinsics in ASan - emit calls into runtime library that poison memory for local variables when their lifetime is over and unpoison memory when their lifetime begins. llvm-svn: 169200
2012-12-04 09:34:23 +08:00
entry:
%i = alloca i32, align 4
%i.ptr = bitcast i32* %i to i8*
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; Poison memory in prologue: F1F1F1F104F3F3F3
; CHECK: store i64 -868083100587789839, i64* %{{[0-9]+}}
Add address space mangling to lifetime intrinsics In preparation for allowing allocas to have non-0 addrspace. llvm-svn: 299876
2017-04-11 04:18:21 +08:00
call void @llvm.lifetime.start.p0i8(i64 -1, i8* %i.ptr)
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; Check that lifetime with no size are ignored.
; CHECK-NOT: store
; CHECK: call void @llvm.lifetime.start
[asan] Skip promotable allocas to improve performance at -O0 Currently, the ASan executables built with -O0 are unnecessarily slow. The main reason is that ASan instrumentation pass inserts redundant checks around promotable allocas. These allocas do not get instrumented under -O1 because they get converted to virtual registered by mem2reg. With this patch, ASan instrumentation pass will only instrument non promotable allocas, giving us a speedup of 39% on a collection of benchmarks with -O0. (There is no measurable speedup at -O1.) llvm-svn: 230724
2015-02-27 11:12:36 +08:00
store volatile i8 0, i8* %i.ptr
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; CHECK: store volatile
Add address space mangling to lifetime intrinsics In preparation for allowing allocas to have non-0 addrspace. llvm-svn: 299876
2017-04-11 04:18:21 +08:00
call void @llvm.lifetime.end.p0i8(i64 -1, i8* %i.ptr)
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; Check that lifetime with no size are ignored.
; CHECK-NOT: store
; CHECK: call void @llvm.lifetime.end
ASan: add initial support for handling llvm.lifetime intrinsics in ASan - emit calls into runtime library that poison memory for local variables when their lifetime is over and unpoison memory when their lifetime begins. llvm-svn: 169200
2012-12-04 09:34:23 +08:00
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; Unpoison stack frame on exit.
; CHECK: store i64 0, i64* %{{[0-9]+}}
; CHECK: ret void
ASan: add initial support for handling llvm.lifetime intrinsics in ASan - emit calls into runtime library that poison memory for local variables when their lifetime is over and unpoison memory when their lifetime begins. llvm-svn: 169200
2012-12-04 09:34:23 +08:00
ret void
}
; Generic case of lifetime analysis.
Unify clang/llvm attributes for asan/tsan/msan (LLVM part) These are two related changes (one in llvm, one in clang). LLVM: - rename address_safety => sanitize_address (the enum value is the same, so we preserve binary compatibility with old bitcode) - rename thread_safety => sanitize_thread - rename no_uninitialized_checks -> sanitize_memory CLANG: - add __attribute__((no_sanitize_address)) as a synonym for __attribute__((no_address_safety_analysis)) - add __attribute__((no_sanitize_thread)) - add __attribute__((no_sanitize_memory)) for S in address thread memory If -fsanitize=S is present and __attribute__((no_sanitize_S)) is not set llvm attribute sanitize_S llvm-svn: 176075
2013-02-26 14:58:09 +08:00
define void @lifetime() sanitize_address {
[ASan] Fix PR17867 - make sure ASan doesn't crash if use-after-scope and use-after-return are combined. llvm-svn: 195014
2013-11-18 22:53:55 +08:00
; CHECK-LABEL: define void @lifetime()
ASan: add initial support for handling llvm.lifetime intrinsics in ASan - emit calls into runtime library that poison memory for local variables when their lifetime is over and unpoison memory when their lifetime begins. llvm-svn: 169200
2012-12-04 09:34:23 +08:00
; Regular variable lifetime intrinsics.
%i = alloca i32, align 4
%i.ptr = bitcast i32* %i to i8*
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; Poison memory in prologue: F1F1F1F1F8F3F3F3
; CHECK: store i64 -868082052615769615, i64* %{{[0-9]+}}
; Memory is unpoisoned at llvm.lifetime.start
Add address space mangling to lifetime intrinsics In preparation for allowing allocas to have non-0 addrspace. llvm-svn: 299876
2017-04-11 04:18:21 +08:00
call void @llvm.lifetime.start.p0i8(i64 3, i8* %i.ptr)
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; CHECK: store i8 4, i8* %{{[0-9]+}}
; CHECK-NEXT: llvm.lifetime.start
[asan] Skip promotable allocas to improve performance at -O0 Currently, the ASan executables built with -O0 are unnecessarily slow. The main reason is that ASan instrumentation pass inserts redundant checks around promotable allocas. These allocas do not get instrumented under -O1 because they get converted to virtual registered by mem2reg. With this patch, ASan instrumentation pass will only instrument non promotable allocas, giving us a speedup of 39% on a collection of benchmarks with -O0. (There is no measurable speedup at -O1.) llvm-svn: 230724
2015-02-27 11:12:36 +08:00
store volatile i8 0, i8* %i.ptr
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; CHECK: store volatile
Add address space mangling to lifetime intrinsics In preparation for allowing allocas to have non-0 addrspace. llvm-svn: 299876
2017-04-11 04:18:21 +08:00
call void @llvm.lifetime.end.p0i8(i64 4, i8* %i.ptr)
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; CHECK: store i8 -8, i8* %{{[0-9]+}}
; CHECK-NEXT: call void @llvm.lifetime.end
ASan: add initial support for handling llvm.lifetime intrinsics in ASan - emit calls into runtime library that poison memory for local variables when their lifetime is over and unpoison memory when their lifetime begins. llvm-svn: 169200
2012-12-04 09:34:23 +08:00
; Memory is poisoned at every call to llvm.lifetime.end
Add address space mangling to lifetime intrinsics In preparation for allowing allocas to have non-0 addrspace. llvm-svn: 299876
2017-04-11 04:18:21 +08:00
call void @llvm.lifetime.end.p0i8(i64 2, i8* %i.ptr)
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; CHECK: store i8 -8, i8* %{{[0-9]+}}
; CHECK-NEXT: call void @llvm.lifetime.end
ASan: add initial support for handling llvm.lifetime intrinsics in ASan - emit calls into runtime library that poison memory for local variables when their lifetime is over and unpoison memory when their lifetime begins. llvm-svn: 169200
2012-12-04 09:34:23 +08:00
; Lifetime intrinsics for array.
%arr = alloca [10 x i32], align 16
%arr.ptr = bitcast [10 x i32]* %arr to i8*
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
Add address space mangling to lifetime intrinsics In preparation for allowing allocas to have non-0 addrspace. llvm-svn: 299876
2017-04-11 04:18:21 +08:00
call void @llvm.lifetime.start.p0i8(i64 40, i8* %arr.ptr)
Revert "[asan] Avoid lifetime analysis for allocas with can be in ambiguous state" This approach is not good enough. Working on the new solution. This reverts commit r280907. llvm-svn: 281689
2016-09-16 09:38:46 +08:00
; CHECK-DEFAULT: call void @__asan_unpoison_stack_memory(i64 %{{[^ ]+}}, i64 40)
Make sure that not interesting allocas are not instrumented. Summary: We failed to unpoison uninteresting allocas on return as unpoisoning is part of main instrumentation which skips such allocas. Added check -asan-instrument-allocas for dynamic allocas. If instrumentation of dynamic allocas is disabled it will not will not be unpoisoned. PR27453 Reviewers: kcc, eugenis Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D21207 llvm-svn: 272341
2016-06-10 07:31:59 +08:00
; CHECK-NO-DYNAMIC-NOT: call void @__asan_unpoison_stack_memory(i64 %{{[^ ]+}}, i64 40)
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
store volatile i8 0, i8* %arr.ptr
; CHECK: store volatile
Add address space mangling to lifetime intrinsics In preparation for allowing allocas to have non-0 addrspace. llvm-svn: 299876
2017-04-11 04:18:21 +08:00
call void @llvm.lifetime.end.p0i8(i64 40, i8* %arr.ptr)
Revert "[asan] Avoid lifetime analysis for allocas with can be in ambiguous state" This approach is not good enough. Working on the new solution. This reverts commit r280907. llvm-svn: 281689
2016-09-16 09:38:46 +08:00
; CHECK-DEFAULT: call void @__asan_poison_stack_memory(i64 %{{[^ ]+}}, i64 40)
Make sure that not interesting allocas are not instrumented. Summary: We failed to unpoison uninteresting allocas on return as unpoisoning is part of main instrumentation which skips such allocas. Added check -asan-instrument-allocas for dynamic allocas. If instrumentation of dynamic allocas is disabled it will not will not be unpoisoned. PR27453 Reviewers: kcc, eugenis Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D21207 llvm-svn: 272341
2016-06-10 07:31:59 +08:00
; CHECK-NO-DYNAMIC-NOT: call void @__asan_poison_stack_memory(i64 %{{[^ ]+}}, i64 40)
ASan: add initial support for handling llvm.lifetime intrinsics in ASan - emit calls into runtime library that poison memory for local variables when their lifetime is over and unpoison memory when their lifetime begins. llvm-svn: 169200
2012-12-04 09:34:23 +08:00
; One more lifetime start/end for the same variable %i.
Add address space mangling to lifetime intrinsics In preparation for allowing allocas to have non-0 addrspace. llvm-svn: 299876
2017-04-11 04:18:21 +08:00
call void @llvm.lifetime.start.p0i8(i64 2, i8* %i.ptr)
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; CHECK: store i8 4, i8* %{{[0-9]+}}
; CHECK-NEXT: llvm.lifetime.start
[asan] Skip promotable allocas to improve performance at -O0 Currently, the ASan executables built with -O0 are unnecessarily slow. The main reason is that ASan instrumentation pass inserts redundant checks around promotable allocas. These allocas do not get instrumented under -O1 because they get converted to virtual registered by mem2reg. With this patch, ASan instrumentation pass will only instrument non promotable allocas, giving us a speedup of 39% on a collection of benchmarks with -O0. (There is no measurable speedup at -O1.) llvm-svn: 230724
2015-02-27 11:12:36 +08:00
store volatile i8 0, i8* %i.ptr
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; CHECK: store volatile
Add address space mangling to lifetime intrinsics In preparation for allowing allocas to have non-0 addrspace. llvm-svn: 299876
2017-04-11 04:18:21 +08:00
call void @llvm.lifetime.end.p0i8(i64 4, i8* %i.ptr)
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; CHECK: store i8 -8, i8* %{{[0-9]+}}
; CHECK-NEXT: llvm.lifetime.end
ASan: add initial support for handling llvm.lifetime intrinsics in ASan - emit calls into runtime library that poison memory for local variables when their lifetime is over and unpoison memory when their lifetime begins. llvm-svn: 169200
2012-12-04 09:34:23 +08:00
; Memory is unpoisoned at function exit (only once).
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; CHECK: store i64 0, i64* %{{[0-9]+}}
; CHECK-NEXT: ret void
ASan: add initial support for handling llvm.lifetime intrinsics in ASan - emit calls into runtime library that poison memory for local variables when their lifetime is over and unpoison memory when their lifetime begins. llvm-svn: 169200
2012-12-04 09:34:23 +08:00
ret void
}
[ASan] Fix lifetime intrinsics handling. Now for each intrinsic we check if it describes one of 'interesting' allocas. Assume that allocas can go through casts and phi-nodes before apperaring as llvm.lifetime arguments llvm-svn: 171153
2012-12-27 16:50:58 +08:00
; Check that arguments of lifetime may come from phi nodes.
Unify clang/llvm attributes for asan/tsan/msan (LLVM part) These are two related changes (one in llvm, one in clang). LLVM: - rename address_safety => sanitize_address (the enum value is the same, so we preserve binary compatibility with old bitcode) - rename thread_safety => sanitize_thread - rename no_uninitialized_checks -> sanitize_memory CLANG: - add __attribute__((no_sanitize_address)) as a synonym for __attribute__((no_address_safety_analysis)) - add __attribute__((no_sanitize_thread)) - add __attribute__((no_sanitize_memory)) for S in address thread memory If -fsanitize=S is present and __attribute__((no_sanitize_S)) is not set llvm attribute sanitize_S llvm-svn: 176075
2013-02-26 14:58:09 +08:00
define void @phi_args(i1 %x) sanitize_address {
[ASan] Fix PR17867 - make sure ASan doesn't crash if use-after-scope and use-after-return are combined. llvm-svn: 195014
2013-11-18 22:53:55 +08:00
; CHECK-LABEL: define void @phi_args(i1 %x)
[ASan] Fix lifetime intrinsics handling. Now for each intrinsic we check if it describes one of 'interesting' allocas. Assume that allocas can go through casts and phi-nodes before apperaring as llvm.lifetime arguments llvm-svn: 171153
2012-12-27 16:50:58 +08:00
entry:
%i = alloca i64, align 4
%i.ptr = bitcast i64* %i to i8*
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; Poison memory in prologue: F1F1F1F1F8F3F3F3
; CHECK: store i64 -868082052615769615, i64* %{{[0-9]+}}
Add address space mangling to lifetime intrinsics In preparation for allowing allocas to have non-0 addrspace. llvm-svn: 299876
2017-04-11 04:18:21 +08:00
call void @llvm.lifetime.start.p0i8(i64 8, i8* %i.ptr)
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; CHECK: store i8 0, i8* %{{[0-9]+}}
; CHECK-NEXT: llvm.lifetime.start
[asan] Skip promotable allocas to improve performance at -O0 Currently, the ASan executables built with -O0 are unnecessarily slow. The main reason is that ASan instrumentation pass inserts redundant checks around promotable allocas. These allocas do not get instrumented under -O1 because they get converted to virtual registered by mem2reg. With this patch, ASan instrumentation pass will only instrument non promotable allocas, giving us a speedup of 39% on a collection of benchmarks with -O0. (There is no measurable speedup at -O1.) llvm-svn: 230724
2015-02-27 11:12:36 +08:00
store volatile i8 0, i8* %i.ptr
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; CHECK: store volatile
[ASan] Fix lifetime intrinsics handling. Now for each intrinsic we check if it describes one of 'interesting' allocas. Assume that allocas can go through casts and phi-nodes before apperaring as llvm.lifetime arguments llvm-svn: 171153
2012-12-27 16:50:58 +08:00
br i1 %x, label %bb0, label %bb1
bb0:
%i.ptr2 = bitcast i64* %i to i8*
br label %bb1
bb1:
%i.phi = phi i8* [ %i.ptr, %entry ], [ %i.ptr2, %bb0 ]
Add address space mangling to lifetime intrinsics In preparation for allowing allocas to have non-0 addrspace. llvm-svn: 299876
2017-04-11 04:18:21 +08:00
call void @llvm.lifetime.end.p0i8(i64 8, i8* %i.phi)
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; CHECK: store i8 -8, i8* %{{[0-9]+}}
; CHECK-NEXT: llvm.lifetime.end
[ASan] Fix lifetime intrinsics handling. Now for each intrinsic we check if it describes one of 'interesting' allocas. Assume that allocas can go through casts and phi-nodes before apperaring as llvm.lifetime arguments llvm-svn: 171153
2012-12-27 16:50:58 +08:00
ret void
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; CHECK: store i64 0, i64* %{{[0-9]+}}
; CHECK-NEXT: ret void
[ASan] Fix lifetime intrinsics handling. Now for each intrinsic we check if it describes one of 'interesting' allocas. Assume that allocas can go through casts and phi-nodes before apperaring as llvm.lifetime arguments llvm-svn: 171153
2012-12-27 16:50:58 +08:00
}
Make sure that not interesting allocas are not instrumented. Summary: We failed to unpoison uninteresting allocas on return as unpoisoning is part of main instrumentation which skips such allocas. Added check -asan-instrument-allocas for dynamic allocas. If instrumentation of dynamic allocas is disabled it will not will not be unpoisoned. PR27453 Reviewers: kcc, eugenis Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D21207 llvm-svn: 272341
2016-06-10 07:31:59 +08:00
Fix detection of stack-use-after scope for char arrays. Summary: Clang inserts GetElementPtrInst so findAllocaForValue was not able to find allocas. PR27453 Reviewers: kcc, eugenis Differential Revision: https://reviews.llvm.org/D22657 llvm-svn: 276374
2016-07-22 08:56:17 +08:00
; Check that arguments of lifetime may come from getelementptr nodes.
define void @getelementptr_args() sanitize_address{
; CHECK-LABEL: define void @getelementptr_args
entry:
%x = alloca [1024 x i8], align 16
%d = alloca i8*, align 8
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; F1F1F1F1
; CHECK: store i32 -235802127, i32* %{{[0-9]+}}
; F3F3F3F3F3F3F3F3
; CHECK: store i64 -868082074056920077, i64* %{{[0-9]+}}
; F3F3F3F3F3F3F3F3
; CHECK: store i64 -868082074056920077, i64* %{{[0-9]+}}
Fix detection of stack-use-after scope for char arrays. Summary: Clang inserts GetElementPtrInst so findAllocaForValue was not able to find allocas. PR27453 Reviewers: kcc, eugenis Differential Revision: https://reviews.llvm.org/D22657 llvm-svn: 276374
2016-07-22 08:56:17 +08:00
%0 = getelementptr inbounds [1024 x i8], [1024 x i8]* %x, i64 0, i64 0
Add address space mangling to lifetime intrinsics In preparation for allowing allocas to have non-0 addrspace. llvm-svn: 299876
2017-04-11 04:18:21 +08:00
call void @llvm.lifetime.start.p0i8(i64 1024, i8* %0)
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; CHECK: call void @__asan_set_shadow_00(i64 %{{[0-9]+}}, i64 128)
; CHECK-NEXT: call void @llvm.lifetime.start
Fix detection of stack-use-after scope for char arrays. Summary: Clang inserts GetElementPtrInst so findAllocaForValue was not able to find allocas. PR27453 Reviewers: kcc, eugenis Differential Revision: https://reviews.llvm.org/D22657 llvm-svn: 276374
2016-07-22 08:56:17 +08:00
store i8* %0, i8** %d, align 8
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; CHECK: store i8
Fix detection of stack-use-after scope for char arrays. Summary: Clang inserts GetElementPtrInst so findAllocaForValue was not able to find allocas. PR27453 Reviewers: kcc, eugenis Differential Revision: https://reviews.llvm.org/D22657 llvm-svn: 276374
2016-07-22 08:56:17 +08:00
Add address space mangling to lifetime intrinsics In preparation for allowing allocas to have non-0 addrspace. llvm-svn: 299876
2017-04-11 04:18:21 +08:00
call void @llvm.lifetime.end.p0i8(i64 1024, i8* %0)
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; CHECK: call void @__asan_set_shadow_f8(i64 %{{[0-9]+}}, i64 128)
; CHECK-NEXT: call void @llvm.lifetime.end
Fix detection of stack-use-after scope for char arrays. Summary: Clang inserts GetElementPtrInst so findAllocaForValue was not able to find allocas. PR27453 Reviewers: kcc, eugenis Differential Revision: https://reviews.llvm.org/D22657 llvm-svn: 276374
2016-07-22 08:56:17 +08:00
ret void
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; CHECK: call void @__asan_set_shadow_00(i64 %{{[0-9]+}}, i64 148)
; CHECK-NEXT: ret void
Fix detection of stack-use-after scope for char arrays. Summary: Clang inserts GetElementPtrInst so findAllocaForValue was not able to find allocas. PR27453 Reviewers: kcc, eugenis Differential Revision: https://reviews.llvm.org/D22657 llvm-svn: 276374
2016-07-22 08:56:17 +08:00
}
Make sure that not interesting allocas are not instrumented. Summary: We failed to unpoison uninteresting allocas on return as unpoisoning is part of main instrumentation which skips such allocas. Added check -asan-instrument-allocas for dynamic allocas. If instrumentation of dynamic allocas is disabled it will not will not be unpoisoned. PR27453 Reviewers: kcc, eugenis Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D21207 llvm-svn: 272341
2016-06-10 07:31:59 +08:00
define void @zero_sized(i64 %a) #0 {
; CHECK-LABEL: define void @zero_sized(i64 %a)
entry:
%a.addr = alloca i64, align 8
%b = alloca [0 x i8], align 1
store i64 %a, i64* %a.addr, align 8
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
Make sure that not interesting allocas are not instrumented. Summary: We failed to unpoison uninteresting allocas on return as unpoisoning is part of main instrumentation which skips such allocas. Added check -asan-instrument-allocas for dynamic allocas. If instrumentation of dynamic allocas is disabled it will not will not be unpoisoned. PR27453 Reviewers: kcc, eugenis Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D21207 llvm-svn: 272341
2016-06-10 07:31:59 +08:00
%0 = bitcast [0 x i8]* %b to i8*
Add address space mangling to lifetime intrinsics In preparation for allowing allocas to have non-0 addrspace. llvm-svn: 299876
2017-04-11 04:18:21 +08:00
call void @llvm.lifetime.start.p0i8(i64 0, i8* %0) #2
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; CHECK: %{{[0-9]+}} = bitcast
; CHECK-NEXT: call void @llvm.lifetime.start
Make sure that not interesting allocas are not instrumented. Summary: We failed to unpoison uninteresting allocas on return as unpoisoning is part of main instrumentation which skips such allocas. Added check -asan-instrument-allocas for dynamic allocas. If instrumentation of dynamic allocas is disabled it will not will not be unpoisoned. PR27453 Reviewers: kcc, eugenis Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D21207 llvm-svn: 272341
2016-06-10 07:31:59 +08:00
%1 = bitcast [0 x i8]* %b to i8*
Add address space mangling to lifetime intrinsics In preparation for allowing allocas to have non-0 addrspace. llvm-svn: 299876
2017-04-11 04:18:21 +08:00
call void @llvm.lifetime.end.p0i8(i64 0, i8* %1) #2
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; CHECK-NEXT: %{{[0-9]+}} = bitcast
; CHECK-NEXT: call void @llvm.lifetime.end
Make sure that not interesting allocas are not instrumented. Summary: We failed to unpoison uninteresting allocas on return as unpoisoning is part of main instrumentation which skips such allocas. Added check -asan-instrument-allocas for dynamic allocas. If instrumentation of dynamic allocas is disabled it will not will not be unpoisoned. PR27453 Reviewers: kcc, eugenis Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D21207 llvm-svn: 272341
2016-06-10 07:31:59 +08:00
ret void
[asan] Enable new stack poisoning with store instruction by default Reviewers: eugenis Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D23968 llvm-svn: 279993
2016-08-30 03:28:34 +08:00
; CHECK-NEXT: ret void
Make sure that not interesting allocas are not instrumented. Summary: We failed to unpoison uninteresting allocas on return as unpoisoning is part of main instrumentation which skips such allocas. Added check -asan-instrument-allocas for dynamic allocas. If instrumentation of dynamic allocas is disabled it will not will not be unpoisoned. PR27453 Reviewers: kcc, eugenis Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D21207 llvm-svn: 272341
2016-06-10 07:31:59 +08:00
}