This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: Trunk frozen starting tomorrow! No new features!
- From: "Joseph S. Myers" <joseph at codesourcery dot com>
- To: Adhemerval Zanella <azanella at linux dot vnet dot ibm dot com>
- Cc: libc-alpha at sourceware dot org
- Date: Fri, 8 Jun 2012 11:09:40 +0000 (UTC)
- Subject: Re: Trunk frozen starting tomorrow! No new features!
- References: <4FCC0B3B.3000006@mentor.com> <4FCCA2FD.9040009@linux.vnet.ibm.com>
On Mon, 4 Jun 2012, Adhemerval Zanella wrote:
> I'm also seeing some unexpected underflows exceptions generated for
> csqrt, j0, j1, y0 and y1 also for ldbl-128ibm, but I don't think I will
> be able to fix them in time for 2.16.
I'd advise adding appropriate UNDERFLOW_EXCEPTION_OK, with comments
referencing relevant bugs (existing or new), in that case.
I believe I understand the possible causes of spurious underflow in csqrt
(bug 14157), that on x86/x86_64 only appear for "double", but fixing them
(for all three implementations) will involve also adding new testcases for
some spurious underflow cases not currently covered by the testsuite - and
those cases may require ulps updates for all architectures, so meaning
they are best avoided during the freeze.
I do not know the causes of spurious underflows from Bessel functions (bug
14155), but I think that bug could reasonably be considered to cover the
long double cases as well.
--
Joseph S. Myers
joseph@codesourcery.com