Re: [PATCH v2] fix ___SYNC () build error when PROFILE_ALL_BRANCHES is enabled

[Thomas Bogendoerfer <tsbogend@alpha.franken.de>]

On Fri, Apr 28, 2023 at 03:13:32PM +0800, genjian zhang wrote:
> On Thu, Apr 27, 2023 at 11:05 PM Genjian <zhanggenjian123@gmail.com> wrote:
> >
> > From: Genjian Zhang <zhanggenjian@kylinos.cn>
> >
> > compiler error (mips-linux-gnu-gcc (Debian 10.2.1-6) 10.2.1 20210110):
> >
> > {standard input}: Assembler messages:
> > {standard input}:171: Error: found '(', expected: ')'
> > {standard input}:171: Error: found '(', expected: ')'
> > {standard input}:171: Error: non-constant expression in ".if" statement
> > {standard input}:171: Error: junk at end of line, first unrecognized
> > character is `('
> >
> > Expands ___SYNC() macros. However, 'if' statement will be wrong
> > assembly.This Compilers report a lot of errors like the above.
> > this problem is caused by the #define of if() in
> > include/linux/compiler.h when CONFIG_PROFILE_ALL_BRANCHES
> > is set. Move '.if' into quoted strings to fix it.
> >
> > Reported-by: k2ci <kernel-bot@kylinos.cn>
> > Signed-off-by: Genjian Zhang <zhanggenjian@kylinos.cn>
> > ---
> >  arch/mips/include/asm/sync.h | 7 ++++---
> >  1 file changed, 4 insertions(+), 3 deletions(-)
> >
> > diff --git a/arch/mips/include/asm/sync.h b/arch/mips/include/asm/sync.h
> > index aabd097933fe..56a4f6f6cdef 100644
> > --- a/arch/mips/include/asm/sync.h
> > +++ b/arch/mips/include/asm/sync.h
> > @@ -175,7 +175,7 @@
> >   */
> >  #ifdef CONFIG_CPU_HAS_SYNC
> >  # define ____SYNC(_type, _reason, _else)                       \
> > -       .if     (( _type ) != -1) && ( _reason );               \
> > +       ((_type) != -1) && (_reason);                           \
> >         .set    push;                                           \
> >         .set    MIPS_ISA_LEVEL_RAW;                             \
> >         .rept   __SYNC_rpt(_type);                              \
> > @@ -192,13 +192,14 @@
> >  /*
> >   * Preprocessor magic to expand macros used as arguments before we insert them
> >   * into assembly code.
> > + * In addition,'if' cannot be enabled.
> >   */
> >  #ifdef __ASSEMBLY__
> >  # define ___SYNC(type, reason, else)                           \
> > -       ____SYNC(type, reason, else)
> > +       .if     ____SYNC(type, reason, else)
> >  #else
> >  # define ___SYNC(type, reason, else)                           \
> > -       __stringify(____SYNC(type, reason, else))
> > +       ".if"   __stringify(____SYNC(type, reason, else))
> >  #endif
> >
> >  #define __SYNC(type, reason)                                   \
> > --
> > 2.25.1
> >
> 
> This patch does not take into account the situation that CPU_HAS_SYNC
> is not set. I will revise it later.

please post as a new mail not as a replay to this thread.

Thomas.

-- 
Crap can work. Given enough thrust pigs will fly, but it's not necessarily a
good idea.                                                [ RFC1925, 2.3 ]