https://llvm.org/bugs/show_bug.cgi?id=23240
Bug ID: 23240
Summary: Less efficient stack adjustment code with -g
Product: libraries
Version: trunk
Hardware: PC
OS: Linux
Status: NEW
Severity: normal
Priority: P
Component: Backend: X86
Assignee: [email protected]
Reporter: [email protected]
CC: [email protected]
Classification: Unclassified
Enabling debug info results in larger code at address 0x13. Without -g a 1 byte
pop instruction is used, with -g a 4 byte add is used to adjust the stack
pointer. Guessing that this is in the X86 backend.
Reduced from Burg in the llvm-test-suite. With revision r234984.
$ cat lex.c
int a;
int fn2();
int fn1() {
int b;
if ((b = fn2()) == '\n')
a++;
return b;
}
$ clang -c -O1 lex.c -o lex.o && objdump -d lex.o
lex.o: file format elf64-x86-64
Disassembly of section .text:
0000000000000000 <fn1>:
0: 50 push %rax
1: 31 c0 xor %eax,%eax
3: e8 00 00 00 00 callq 8 <fn1+0x8>
8: 83 f8 0a cmp $0xa,%eax
b: 75 06 jne 13 <fn1+0x13>
d: ff 05 00 00 00 00 incl 0x0(%rip) # 13 <fn1+0x13>
13: 5a pop %rdx
14: c3 retq
$ clang -c -O1 lex.c -o lex_g.o -g && objdump -d lex_g.o
lex_g.o: file format elf64-x86-64
Disassembly of section .text:
0000000000000000 <fn1>:
0: 50 push %rax
1: 31 c0 xor %eax,%eax
3: e8 00 00 00 00 callq 8 <fn1+0x8>
8: 83 f8 0a cmp $0xa,%eax
b: 75 06 jne 13 <fn1+0x13>
d: ff 05 00 00 00 00 incl 0x0(%rip) # 13 <fn1+0x13>
13: 48 83 c4 08 add $0x8,%rsp
17: c3 retq
--
You are receiving this mail because:
You are on the CC list for the bug.
_______________________________________________
LLVMbugs mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/llvmbugs
