[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: libc build failure
From: |
Igor Khavkine |
Subject: |
Re: libc build failure |
Date: |
Tue, 21 Aug 2001 15:52:32 -0400 |
User-agent: |
Mutt/1.3.20i |
On Tue, Aug 21, 2001 at 03:37:32PM -0400, Roland McGrath wrote:
> I can explain that if and only if you compiled libc without optimization.
> The various `extern inline' functions are supposed to have real-function
> versions defined for unoptimized callers. But __hurd_fail was missing.
> I've just now fixed that in libc. For __thread_stack_pointer, there is
> no such real-function definition. I am inclined to just make that one a
> macro.
It seems to me that if using gcc and replacing `extern inline' with
`static inline' this problem would go away. The gcc docs say that
these functions will be inline, but there will be a normally compiled
version as well if there is need (for example the function's pointer
was used somewhere). There are also questions about C99 and future
standards compatibility where `static inline' seems to be favored.
I've raised the `extern inline' vs `static inline' question on the
glibc mailing list before. But no-one seemed interested in discussing it.
Igor
- libc build failure, Neal H Walfield, 2001/08/20
- Re: libc build failure, Roland McGrath, 2001/08/20
- Re: libc build failure, Neal H Walfield, 2001/08/21
- Re: libc build failure, Roland McGrath, 2001/08/21
- Re: libc build failure,
Igor Khavkine <=
- Re: libc build failure, Roland McGrath, 2001/08/21
- Re: libc build failure, Igor Khavkine, 2001/08/21
- Re: libc build failure, Roland McGrath, 2001/08/21
- Re: libc build failure, Igor Khavkine, 2001/08/21