caio/clang-p2996
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
c6c7bfc4d2bc37f55f57504c19d47deb7645dd76
clang-p2996/llvm/test/CodeGen/X86/win32_sret.ll
Reid Kleckner 3a7a2e4a0a [FastISel] Sink local value materializations to first use 
Summary:
Local values are constants, global addresses, and stack addresses that
can't be folded into the instruction that uses them. For example, when
storing the address of a global variable into memory, we need to
materialize that address into a register.

FastISel doesn't want to materialize any given local value more than
once, so it generates all local value materialization code at
EmitStartPt, which always dominates the current insertion point. This
allows it to maintain a map of local value registers, and it knows that
the local value area will always dominate the current insertion point.

The downside is that local value instructions are always emitted without
a source location. This is done to prevent jumpy line tables, but it
means that the local value area will be considered part of the previous
statement. Consider this C code:
  call1();      // line 1
  ++global;     // line 2
  ++global;     // line 3
  call2(&global, &local); // line 4

Today we end up with assembly and line tables like this:
  .loc 1 1
  callq call1
  leaq global(%rip), %rdi
  leaq local(%rsp), %rsi
  .loc 1 2
  addq $1, global(%rip)
  .loc 1 3
  addq $1, global(%rip)
  .loc 1 4
  callq call2

The LEA instructions in the local value area have no source location and
are treated as being on line 1. Stepping through the code in a debugger
and correlating it with the assembly won't make much sense, because
these materializations are only required for line 4.

This is actually problematic for the VS debugger "set next statement"
feature, which effectively assumes that there are no registers live
across statement boundaries. By sinking the local value code into the
statement and fixing up the source location, we can make that feature
work. This was filed as https://bugs.llvm.org/show_bug.cgi?id=35975 and
https://crbug.com/793819.

This change is obviously not enough to make this feature work reliably
in all cases, but I felt that it was worth doing anyway because it
usually generates smaller, more comprehensible -O0 code. I measured a
0.12% regression in code generation time with LLC on the sqlite3
amalgamation, so I think this is worth doing.

There are some special cases worth calling out in the commit message:
1. local values materialized for phis
2. local values used by no-op casts
3. dead local value code

Local values can be materialized for phis, and this does not show up as
a vreg use in MachineRegisterInfo. In this case, if there are no other
uses, this patch sinks the value to the first terminator, EH label, or
the end of the BB if nothing else exists.

Local values may also be used by no-op casts, which adds the register to
the RegFixups table. Without reversing the RegFixups map direction, we
don't have enough information to sink these instructions.

Lastly, if the local value register has no other uses, we can delete it.
This comes up when fastisel tries two instruction selection approaches
and the first materializes the value but fails and the second succeeds
without using the local value.

Reviewers: aprantl, dblaikie, qcolombet, MatzeB, vsk, echristo

Subscribers: dotdash, chandlerc, hans, sdardis, amccarth, javed.absar, zturner, llvm-commits, hiraditya

Differential Revision: https://reviews.llvm.org/D43093

llvm-svn: 327581
2018-03-14 21:54:21 +00:00

239 lines
7.1 KiB
LLVM
Raw Blame History

; We specify -mcpu explicitly to avoid instruction reordering that happens on
; some setups (e.g., Atom) from affecting the output.
; RUN: llc < %s -mcpu=core2 -mtriple=i686-pc-win32 | FileCheck %s -check-prefix=WIN32
; RUN: llc < %s -mcpu=core2 -mtriple=i686-pc-mingw32 | FileCheck %s -check-prefix=MINGW_X86
; RUN: llc < %s -mcpu=core2 -mtriple=i686-pc-cygwin | FileCheck %s -check-prefix=CYGWIN
; RUN: llc < %s -mcpu=core2 -mtriple=i386-pc-linux | FileCheck %s -check-prefix=LINUX
; RUN: llc < %s -mcpu=core2 -O0 -mtriple=i686-pc-win32 | FileCheck %s -check-prefix=WIN32
; RUN: llc < %s -mcpu=core2 -O0 -mtriple=i686-pc-mingw32 | FileCheck %s -check-prefix=MINGW_X86
; RUN: llc < %s -mcpu=core2 -O0 -mtriple=i686-pc-cygwin | FileCheck %s -check-prefix=CYGWIN
; RUN: llc < %s -mcpu=core2 -O0 -mtriple=i386-pc-linux | FileCheck %s -check-prefix=LINUX
; The SysV ABI used by most Unixes and Mingw on x86 specifies that an sret pointer
; is callee-cleanup. However, in MSVC's cdecl calling convention, sret pointer
; arguments are caller-cleanup like normal arguments.
define void @sret1(i8* sret %x) nounwind {
entry:
; WIN32-LABEL: _sret1:
; WIN32: movb $42, (%eax)
; WIN32-NOT: popl %eax
; WIN32: {{retl$}}
; MINGW_X86-LABEL: _sret1:
; MINGW_X86: {{retl$}}
; CYGWIN-LABEL: _sret1:
; CYGWIN: retl $4
; LINUX-LABEL: sret1:
; LINUX: retl $4
store i8 42, i8* %x, align 4
ret void
}
define void @sret2(i8* sret %x, i8 %y) nounwind {
entry:
; WIN32-LABEL: _sret2:
; WIN32: movb {{.*}}, (%eax)
; WIN32-NOT: popl %eax
; WIN32: {{retl$}}
; MINGW_X86-LABEL: _sret2:
; MINGW_X86: {{retl$}}
; CYGWIN-LABEL: _sret2:
; CYGWIN: retl $4
; LINUX-LABEL: sret2:
; LINUX: retl $4
store i8 %y, i8* %x
ret void
}
define void @sret3(i8* sret %x, i8* %y) nounwind {
entry:
; WIN32-LABEL: _sret3:
; WIN32: movb $42, (%eax)
; WIN32-NOT: movb $13, (%eax)
; WIN32-NOT: popl %eax
; WIN32: {{retl$}}
; MINGW_X86-LABEL: _sret3:
; MINGW_X86: {{retl$}}
; CYGWIN-LABEL: _sret3:
; CYGWIN: retl $4
; LINUX-LABEL: sret3:
; LINUX: retl $4
store i8 42, i8* %x
store i8 13, i8* %y
ret void
}
; PR15556
%struct.S4 = type { i32, i32, i32 }
define void @sret4(%struct.S4* noalias sret %agg.result) {
entry:
; WIN32-LABEL: _sret4:
; WIN32: movl $42, (%eax)
; WIN32-NOT: popl %eax
; WIN32: {{retl$}}
; MINGW_X86-LABEL: _sret4:
; MINGW_X86: {{retl$}}
; CYGWIN-LABEL: _sret4:
; CYGWIN: retl $4
; LINUX-LABEL: sret4:
; LINUX: retl $4
%x = getelementptr inbounds %struct.S4, %struct.S4* %agg.result, i32 0, i32 0
store i32 42, i32* %x, align 4
ret void
}
%struct.S5 = type { i32 }
%class.C5 = type { i8 }
define x86_thiscallcc void @"\01?foo@C5@@QAE?AUS5@@XZ"(%struct.S5* noalias sret %agg.result, %class.C5* %this) {
entry:
%this.addr = alloca %class.C5*, align 4
store %class.C5* %this, %class.C5** %this.addr, align 4
%this1 = load %class.C5*, %class.C5** %this.addr
%x = getelementptr inbounds %struct.S5, %struct.S5* %agg.result, i32 0, i32 0
store i32 42, i32* %x, align 4
ret void
; WIN32-LABEL: {{^}}"?foo@C5@@QAE?AUS5@@XZ":
; MINGW_X86-LABEL: {{^}}"?foo@C5@@QAE?AUS5@@XZ":
; CYGWIN-LABEL: {{^}}"?foo@C5@@QAE?AUS5@@XZ":
; LINUX-LABEL: {{^}}"?foo@C5@@QAE?AUS5@@XZ":
; The address of the return structure is passed as an implicit parameter.
; In the -O0 build, %eax is spilled at the beginning of the function, hence we
; should match both 4(%esp) and 8(%esp).
; WIN32: {{[48]}}(%esp), %eax
; WIN32: movl $42, (%eax)
; WIN32: retl $4
}
define void @call_foo5() {
entry:
%c = alloca %class.C5, align 1
%s = alloca %struct.S5, align 4
call x86_thiscallcc void @"\01?foo@C5@@QAE?AUS5@@XZ"(%struct.S5* sret %s, %class.C5* %c)
; WIN32-LABEL: {{^}}_call_foo5:
; MINGW_X86-LABEL: {{^}}_call_foo5:
; CYGWIN-LABEL: {{^}}_call_foo5:
; LINUX-LABEL: {{^}}call_foo5:
; Load the address of the result and put it onto stack
; The this pointer goes to ECX.
; (through %ecx in the -O0 build).
; WIN32-DAG: leal {{[0-9]*}}(%esp), %e{{[a-d]}}x
; WIN32-DAG: {{leal [1-9]+\(%esp\)|movl %esp}}, %ecx
; WIN32-DAG: {{pushl %e[a-d]x|movl %e[a-d]x, \(%esp\)}}
; WIN32-NEXT: calll "?foo@C5@@QAE?AUS5@@XZ"
; WIN32: retl
ret void
}
%struct.test6 = type { i32, i32, i32 }
define void @test6_f(%struct.test6* %x) nounwind {
; WIN32-LABEL: _test6_f:
; MINGW_X86-LABEL: _test6_f:
; CYGWIN-LABEL: _test6_f:
; LINUX-LABEL: test6_f:
; The %x argument is moved to %ecx. It will be the this pointer.
; WIN32-DAG: movl {{16|20}}(%esp), %ecx
; The sret pointer is (%esp)
; WIN32-DAG: {{leal 4\(%esp\)|movl %esp}}, %eax
; WIN32-DAG: {{pushl %eax|movl %eax, \(%esp\)}}
; The sret pointer is %ecx
; The %x argument is moved to (%esp). It will be the this pointer.
; MINGW_X86-DAG: {{leal 4\(%esp\)|movl %esp}}, %ecx
; MINGW_X86-DAG: {{pushl 16\(%esp\)|movl %eax, \(%esp\)}}
; MINGW_X86-NEXT: calll _test6_g
; CYGWIN-DAG: {{leal 4\(%esp\)|movl %esp}}, %ecx
; CYGWIN-DAG: {{pushl 16\(%esp\)|movl %eax, \(%esp\)}}
; CYGWIN-NEXT: calll _test6_g
%tmp = alloca %struct.test6, align 4
call x86_thiscallcc void @test6_g(%struct.test6* sret %tmp, %struct.test6* %x)
ret void
}
declare x86_thiscallcc void @test6_g(%struct.test6* sret, %struct.test6*)
; Flipping the parameters at the IR level generates the same code.
%struct.test7 = type { i32, i32, i32 }
define void @test7_f(%struct.test7* %x) nounwind {
; WIN32-LABEL: _test7_f:
; MINGW_X86-LABEL: _test7_f:
; CYGWIN-LABEL: _test7_f:
; LINUX-LABEL: test7_f:
; The %x argument is moved to %ecx on all OSs. It will be the this pointer.
; WIN32: movl {{16|20}}(%esp), %ecx
; MINGW_X86: movl {{16|20}}(%esp), %ecx
; CYGWIN: movl {{16|20}}(%esp), %ecx
; The sret pointer is (%esp)
; WIN32: {{leal 4\(%esp\)|movl %esp}}, %eax
; WIN32-NEXT: {{pushl %eax|movl %eax, \(%esp\)}}
; MINGW_X86: {{leal 4\(%esp\)|movl %esp}}, %eax
; MINGW_X86-NEXT: {{pushl %eax|movl %eax, \(%esp\)}}
; CYGWIN: {{leal 4\(%esp\)|movl %esp}}, %eax
; CYGWIN-NEXT: {{pushl %eax|movl %eax, \(%esp\)}}
%tmp = alloca %struct.test7, align 4
call x86_thiscallcc void @test7_g(%struct.test7* %x, %struct.test7* sret %tmp)
ret void
}
define x86_thiscallcc void @test7_g(%struct.test7* %in, %struct.test7* sret %out) {
%s = getelementptr %struct.test7, %struct.test7* %in, i32 0, i32 0
%d = getelementptr %struct.test7, %struct.test7* %out, i32 0, i32 0
%v = load i32, i32* %s
store i32 %v, i32* %d
call void @clobber_eax()
ret void
; Make sure we return the second parameter in %eax.
; WIN32-LABEL: _test7_g:
; WIN32: calll _clobber_eax
; WIN32: movl {{.*}}, %eax
; WIN32: retl
}
declare void @clobber_eax()
; Test what happens if the first parameter has to be split by codegen.
; Realistically, no frontend will generate code like this, but here it is for
; completeness.
define void @test8_f(i64 inreg %a, i64* sret %out) {
store i64 %a, i64* %out
call void @clobber_eax()
ret void
; WIN32-LABEL: _test8_f:
; WIN32: movl {{[0-9]+}}(%esp), %[[out:[a-z]+]]
; WIN32-DAG: movl %edx, 4(%[[out]])
; WIN32-DAG: movl %eax, (%[[out]])
; WIN32: calll _clobber_eax
; WIN32: movl {{.*}}, %eax
; WIN32: retl
}
Reference in New Issue View Git Blame Copy Permalink