* Why *l math functions are just skipped for some targets?
@ 2014-10-07 16:44 Sergey Dmitrouk
2014-10-07 17:13 ` Rich Felker
0 siblings, 1 reply; 2+ messages in thread
From: Sergey Dmitrouk @ 2014-10-07 16:44 UTC (permalink / raw)
To: musl
Hi,
files like __fpclassifyl.c omit implementation for corresponding
functions on platforms where sizeof(long double) == sizeof(double).
This is fine when musl is compiled at least with -O1 and compiler
omits calls to non-existing functions, but if one builds debug version
without optimizations, some executables end up having unresolved symbols
for functions like __fpclassifyl().
This causes some inconvenience. Also, optimizing compilers are not
intended to allow such hacks, it just happens to be so. Is there any
reason to not provide some stub functions? I can image small size is
the reason, but not sure it justifies unexpected unresolved symbols. At
least a comment in place of absent functions would be helpful.
Best regards,
Sergey
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: Why *l math functions are just skipped for some targets?
2014-10-07 16:44 Why *l math functions are just skipped for some targets? Sergey Dmitrouk
@ 2014-10-07 17:13 ` Rich Felker
0 siblings, 0 replies; 2+ messages in thread
From: Rich Felker @ 2014-10-07 17:13 UTC (permalink / raw)
To: musl
On Tue, Oct 07, 2014 at 07:44:15PM +0300, Sergey Dmitrouk wrote:
> Hi,
>
> files like __fpclassifyl.c omit implementation for corresponding
> functions on platforms where sizeof(long double) == sizeof(double).
> This is fine when musl is compiled at least with -O1 and compiler
> omits calls to non-existing functions, but if one builds debug version
> without optimizations, some executables end up having unresolved symbols
> for functions like __fpclassifyl().
>
> This causes some inconvenience. Also, optimizing compilers are not
> intended to allow such hacks, it just happens to be so. Is there any
> reason to not provide some stub functions? I can image small size is
> the reason, but not sure it justifies unexpected unresolved symbols. At
> least a comment in place of absent functions would be helpful.
I'm pretty sure it's just an omission that went unnoticed, not any
attempt to optimize for size via hacks. I'll look into getting it
fixed.
Rich
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2014-10-07 17:13 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2014-10-07 16:44 Why *l math functions are just skipped for some targets? Sergey Dmitrouk
2014-10-07 17:13 ` Rich Felker
Code repositories for project(s) associated with this public inbox
https://git.vuxu.org/mirror/musl/
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).