This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PATCH RFC] <math.h> issignaling
On Fri, 15 Mar 2013, Thomas Schwinge wrote:
> following? Instead of intruducing these integer calling conventions, fix
> this issue in GCC (I can't comment on the complexity of this task,
> though), and declare that any unfixed GCC version is not required to do
> the right thing in glibc with respect to sNaNs? (And then XFAIL the
> respective tests for 32-bit x86 until we can depend on such GCC
> versions.)
If that's how you wish to do it. There's no actual XFAIL mechanism here,
so it would be lots of #if conditionals in libm-test.inc (based on a
define from a sysdeps header file, I suppose), though. (The uses of
issignaling to verify that a function return value is a quiet NaN
shouldn't need conditioning, though they might not work properly on 32-bit
x86. And those would cover one purpose of this series - being able to
verify that functions such as sqrtl return the right sort of NaN on MIPS,
where presently they don't.)
> Or perhaps I'm implementing the integer calling conventions differently
> From how you meant it to be done -- did you already have a specific plan
> about the implementation?
To avoid errors about conflicting types you may be able to use asm
("asm-name") on various declarations, to give functions different names in
the compiler output from those that they have at C level.
--
Joseph S. Myers
joseph@codesourcery.com