[llvm-dev] [RFC] __builtin_constant_p() Improvements

[Bill Wendling via llvm-dev]

Hello again!

I took a stab at PR4898[1]. The attached patch improves Clang's
__builtin_constant_p support so that the Linux kernel is happy. With this
improvement, Clang can determine if __builtin_constant_p is true or false
after inlining.

As an example:

static __attribute__((always_inline)) int foo(int x) {
if (__builtin_constant_p(x))
return 1;
return 0;
}

static __attribute__((always_inline)) int mux() {
if (__builtin_constant_p(37))
return 927;
return 0;
}

int bar(int a) {
if (a)
return foo(42);
else
return mux();
}

Now outputs this code at -O1:

bar:
.cfi_startproc
# %bb.0:                                # %entry
testl %edi, %edi
movl $927, %ecx              # imm = 0x39F
movl $1, %eax
cmovel %ecx, %eax
retq

And this code at -O0:

bar:                                    # @bar
.cfi_startproc
# %bb.0:                                # %entry
pushq %rbp
.cfi_def_cfa_offset 16
.cfi_offset %rbp, -16
movq %rsp, %rbp
.cfi_def_cfa_register %rbp
movl %edi, -16(%rbp)
cmpl $0, -16(%rbp)
je .LBB0_2
# %bb.1:                                # %if.then
movl $42, -8(%rbp)
movl $0, -4(%rbp)
movl -4(%rbp), %eax
movl %eax, -12(%rbp)
jmp .LBB0_3
.LBB0_2:                                # %if.else
movl $927, -12(%rbp)         # imm = 0x39F
.LBB0_3:                                # %return
movl -12(%rbp), %eax
popq %rbp
retq

If the patch looks okay to people, I can shove it onto Phabricator for a
review. (My phab-fu is bad.)

Thoughts?

-bw

[1] https://bugs.llvm.org/show_bug.cgi?id=4898
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20180412/8a2d3328/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: bcp.patch
Type: text/x-patch
Size: 4024 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20180412/8a2d3328/attachment.bin>