https://gcc.gnu.org/bugzilla/show_bug.cgi?id=118465
Bug ID: 118465
Summary: un-needed aligning the stack in some cases
Product: gcc
Version: 15.0
Status: UNCONFIRMED
Keywords: missed-optimization
Severity: enhancement
Priority: P3
Component: target
Assignee: unassigned at gcc dot gnu.org
Reporter: pinskia at gcc dot gnu.org
Target Milestone: ---
Target: x86_64
Created attachment 60144
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=60144&action=edit
testcase
At -O3, look at _TAIL_CALL_YIELD_VALUE, there is:
GCC produces:
pushq %rbx
movq -24(%rdi), %rax
addq $2, %rcx
subl $2, %r9d
movq -8(%rsi), %r8
subq $8, %rsi
movb %r9b, -5(%rdi)
movq %rcx, 56(%rdi)
movq %rsi, 64(%rdi)
movq %rax, 120(%rdx)
movq 8(%rdi), %rax
movq $0, -24(%rdi)
movq 56(%rax), %rcx
movq 64(%rax), %rsi
movq %rax, 72(%rdx)
addl $1, 44(%rdx)
movzwl 4(%rcx), %r9d
movq $0, 8(%rdi)
addq $4, %rcx
addq $8, %rsi
movq $0, 64(%rax)
movl %r9d, %ebx
movzbl %r9b, %edi
movq %r8, -8(%rsi)
movzbl %bh, %ebx
movq INSTRUCTION_TABLE(,%rdi,8), %r10
movq %rdi, %r8
movq %rax, %rdi
movl %ebx, %r9d
popq %rbx
jmp *%r10
But rbx is not used nor is the stack. I have attached the original preprocessed
source but I have not extracted a simplified version yet. The push/pop as far
as I can tell is only to align the stack to 16 bytes but it seems not to be
needed as there is no stack usage nor uses of rbx.
Originally reported at https://gcc.gnu.org/bugzilla/show_bug.cgi?id=118328#c7
but that is about preserve_none for aarch64 while the issue being reported in
the comment section is about x86_64.
