Linux 2.6.32.xx build breaks with gcc 4.7

Linux 2.6.32.xx build breaks with gcc 4.7

[Gabor Z. Papp]

lo lo,

I'm trying to compile 2.6.32.60 with gcc 4.7.2, and getting the
following error:

  CC      arch/x86/kernel/ptrace.o
arch/x86/kernel/ptrace.c:1472:17: error: conflicting types for 'syscall_trace_enter'
In file included from /usr/src/linux-2.6.32.60-gzpLinux/arch/x86/include/asm/vm86.h:130:0,
                 from /usr/src/linux-2.6.32.60-gzpLinux/arch/x86/include/asm/processor.h:10,
                 from /usr/src/linux-2.6.32.60-gzpLinux/arch/x86/include/asm/thread_info.h:22,
                 from include/linux/thread_info.h:56,
                 from include/linux/preempt.h:9,
                 from include/linux/spinlock.h:50,
                 from include/linux/seqlock.h:29,
                 from include/linux/time.h:8,
                 from include/linux/timex.h:56,
                 from include/linux/sched.h:56,
                 from arch/x86/kernel/ptrace.c:11:
/usr/src/linux-2.6.32.60-gzpLinux/arch/x86/include/asm/ptrace.h:145:13: note: previous declaration of 'syscall_trace_enter' was here
arch/x86/kernel/ptrace.c:1517:17: error: conflicting types for 'syscall_trace_leave'
In file included from /usr/src/linux-2.6.32.60-gzpLinux/arch/x86/include/asm/vm86.h:130:0,
                 from /usr/src/linux-2.6.32.60-gzpLinux/arch/x86/include/asm/processor.h:10,
                 from /usr/src/linux-2.6.32.60-gzpLinux/arch/x86/include/asm/thread_info.h:22,
                 from include/linux/thread_info.h:56,
                 from include/linux/preempt.h:9,
                 from include/linux/spinlock.h:50,
                 from include/linux/seqlock.h:29,
                 from include/linux/time.h:8,
                 from include/linux/timex.h:56,
                 from include/linux/sched.h:56,
                 from arch/x86/kernel/ptrace.c:11:
/usr/src/linux-2.6.32.60-gzpLinux/arch/x86/include/asm/ptrace.h:146:13: note: previous declaration of 'syscall_trace_leave' was here
make[2]: *** [arch/x86/kernel/ptrace.o] Error 1
make[1]: *** [arch/x86/kernel] Error 2
make: *** [arch/x86] Error 2

My environment looks like:

Linux 3.2.30, glibc 2.16, gcc 4.7.2

If I comment out the ancient line, kernel compiles fine.

Willy says:

| The affected entry is only used there :

| arch/x86/include/asm/ptrace.h:extern void syscall_trace_leave(struct pt_regs *);
| arch/x86/include/asm/thread_info.h:/* work to do in syscall_trace_leave() */
| arch/x86/kernel/entry_32.S:     ENABLE_INTERRUPTS(CLBR_ANY)     # could let syscall_trace_leave() call
| arch/x86/kernel/entry_32.S:     call syscall_trace_leave
| arch/x86/kernel/entry_64.S:     call syscall_trace_leave
| arch/x86/kernel/ptrace.c:asmregparm void syscall_trace_leave(struct pt_regs *regs)

| So if the two calls above don't need the include then maybe we can
| safely remove it.

Re: Linux 2.6.32.xx build breaks with gcc 4.7

[Willy Tarreau]

On Tue, Oct 16, 2012 at 09:32:29AM +0200, Gabor Z. Papp wrote:
> lo lo,
> 
> I'm trying to compile 2.6.32.60 with gcc 4.7.2, and getting the
> following error:
> 
>   CC      arch/x86/kernel/ptrace.o
> arch/x86/kernel/ptrace.c:1472:17: error: conflicting types for 'syscall_trace_enter'
> In file included from /usr/src/linux-2.6.32.60-gzpLinux/arch/x86/include/asm/vm86.h:130:0,
>                  from /usr/src/linux-2.6.32.60-gzpLinux/arch/x86/include/asm/processor.h:10,
>                  from /usr/src/linux-2.6.32.60-gzpLinux/arch/x86/include/asm/thread_info.h:22,
>                  from include/linux/thread_info.h:56,
>                  from include/linux/preempt.h:9,
>                  from include/linux/spinlock.h:50,
>                  from include/linux/seqlock.h:29,
>                  from include/linux/time.h:8,
>                  from include/linux/timex.h:56,
>                  from include/linux/sched.h:56,
>                  from arch/x86/kernel/ptrace.c:11:
> /usr/src/linux-2.6.32.60-gzpLinux/arch/x86/include/asm/ptrace.h:145:13: note: previous declaration of 'syscall_trace_enter' was here
> arch/x86/kernel/ptrace.c:1517:17: error: conflicting types for 'syscall_trace_leave'
> In file included from /usr/src/linux-2.6.32.60-gzpLinux/arch/x86/include/asm/vm86.h:130:0,
>                  from /usr/src/linux-2.6.32.60-gzpLinux/arch/x86/include/asm/processor.h:10,
>                  from /usr/src/linux-2.6.32.60-gzpLinux/arch/x86/include/asm/thread_info.h:22,
>                  from include/linux/thread_info.h:56,
>                  from include/linux/preempt.h:9,
>                  from include/linux/spinlock.h:50,
>                  from include/linux/seqlock.h:29,
>                  from include/linux/time.h:8,
>                  from include/linux/timex.h:56,
>                  from include/linux/sched.h:56,
>                  from arch/x86/kernel/ptrace.c:11:
> /usr/src/linux-2.6.32.60-gzpLinux/arch/x86/include/asm/ptrace.h:146:13: note: previous declaration of 'syscall_trace_leave' was here
> make[2]: *** [arch/x86/kernel/ptrace.o] Error 1
> make[1]: *** [arch/x86/kernel] Error 2
> make: *** [arch/x86] Error 2
> 
> My environment looks like:
> 
> Linux 3.2.30, glibc 2.16, gcc 4.7.2
> 
> If I comment out the ancient line, kernel compiles fine.
> 
> Willy says:
> 
> | The affected entry is only used there :
> 
> | arch/x86/include/asm/ptrace.h:extern void syscall_trace_leave(struct pt_regs *);
> | arch/x86/include/asm/thread_info.h:/* work to do in syscall_trace_leave() */
> | arch/x86/kernel/entry_32.S:     ENABLE_INTERRUPTS(CLBR_ANY)     # could let syscall_trace_leave() call
> | arch/x86/kernel/entry_32.S:     call syscall_trace_leave
> | arch/x86/kernel/entry_64.S:     call syscall_trace_leave
> | arch/x86/kernel/ptrace.c:asmregparm void syscall_trace_leave(struct pt_regs *regs)
> 
> | So if the two calls above don't need the include then maybe we can
> | safely remove it.

If someone knows this area well enough to judge whether we can 
  - either remove the asmregparm (I don't think so judging by recent commits
    saying it was possible to remove it after other main changes)
  - or remove the declaration because noone else needs it,

Then I'd happily apply the fix. From what I understood, Paul would probably
need the same fix for 2.6.34.

Thanks,
Willy