SYS_FUNC macro fix for musl

[Dmitry V. Levin]

On Tue, Dec 15, 2015 at 10:56:49PM +0100, Szabolcs Nagy wrote:
> * Dmitry V. Levin <ldv at altlinux.org> [2015-12-15 14:59:32 +0300]:
> > On Tue, Dec 15, 2015 at 12:02:31PM +0100, Szabolcs Nagy wrote:
> > > * Dmitry V. Levin <ldv at altlinux.org> [2015-12-15 00:35:56 +0300]:
> > > > On Mon, Dec 14, 2015 at 09:05:50PM +0100, Szabolcs Nagy wrote:
> > > > > --- a/linux/ia64/syscallent.h
> > > > > +++ b/linux/ia64/syscallent.h
> > > > > @@ -31,10 +31,10 @@
> > > > >   * with 64-bit layout get redirected to printargs.
> > > > >   */
> > > > >  #undef SYS_FUNC_NAME
> > > > > -#define SYS_FUNC_NAME(syscall_name) printargs
> > > > > +#define SYS_FUNC_NAME(syscall_name) sys_printargs
> > > > 
> > > > You don't need this, sys_printargs is a macro defined to printargs.
> > > 
> > > i haven't tested it on ia64, but it seemed to me
> > > that redirecting sys_* to sys_printargs for i386
> > > syscalls was intentional.
> > 
> > Yes, the redirection itself is intentional, but there is no difference whether
> > it's printargs or sys_printargs.
> 
> ah ok
> 
> attached an updated patch with that fix.

Applied, thanks.

> i ran the tests and had some failures, some of them
> might be musl bugs, but at least the %Lu printf format
> specifier used in some tests is invalid.

No problem, changed all %Ld/%Lu to %lld/%llu.

> (%L is for long double, for long long unsigned use %llu)
> attached the test logs too

Looks like all tests related to %Ld/%Lu failed.
Could you re-run them on v4.10-577-ge67c8e4, please?


-- 
ldv
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 819 bytes
Desc: not available
URL: <http://lists.strace.io/pipermail/strace-devel/attachments/20151216/2d10951f/attachment.bin>