This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [RFC PATCH] Remove bits/string.h.
On 20/06/2017 12:04, Zack Weinberg wrote:
> On Tue, Jun 20, 2017 at 9:54 AM, Adhemerval Zanella
> <adhemerval.zanella@linaro.org> wrote:
>> On 12/06/2017 18:23, Zack Weinberg wrote:
>>> We just got rid of the machine-independent string inlines in the
>>> headers (bits/string2.h) thanks to Wilco; I propose we also eliminate
>>> the machine-dependent ones (bits/string.h). These are not on by
>>> default, and even if they were a good idea at the time they were
>>> introduced, they haven't really been touched in ten to fifteen years
>>> and probably aren't a good idea on current-gen processors.
>>>
>>> With this patch applied, the only effect of __NO_STRING_INLINES is to
>>> disable the __stpcpy -> __builtin_stpcpy and
>>> __mempcpy -> __builtin_mempcpy redirects in include/string.h. I
>>> suspect this can be cleaned up further but I would like to see this
>>> patch stick first.
>>
>> I am seeing multiple build issue due new x86_64 multiarch C implementation
>> after this patch. I am fixing the build and I will post a patch soon.
>
> I believe I have fixed this in
> c0b23001a89b79f8d0bebe41bfbe64d840b13191, please update and try again.
> (Unless you're tackling getting rid of the need for
> NO_MEMPCPY_STPCPY_REDIRECT and __NO_STRING_INLINES, in which case
> carry on ;-)
>
> zw
>
It is fixed now and I think we can cleanup some HAVE_STRING_ARCH_ on x86
code. Thanks for fixed it quickly.