hypotl on OpenBSD 5.1/SPARC

> On OpenBSD 5.1/SPARC64, the sqrtl() function returns values that have a > relative error > 5000 ulp. Obviously unusable. Likewise for the hypotl() function. Witness: #include #include #include volatile long double x; volatile long double y; volatile long double z; int main () { x = 2.5

remainderl on OpenBSD

On OpenBSD 5.1/SPARC64, remainderl() is also completely broken: For x = 9.245907126L, y = 3.141592654L, it returns 0.104314472L, that is, x - y - 6, instead of x - 3*y. Workaround added: 2012-03-13 Bruno Haible remainderl: Bypass broken implementation in OpenBSD 5.1/SPARC

sqrtl on OpenBSD 5.1/SPARC

On OpenBSD 5.1/SPARC64, the sqrtl() function returns values that have a relative error > 5000 ulp. Obviously unusable. Witness: #include #include int main () { volatile long double a = 8.1974099812331540680810141969554806865L; volatile long double b = sqrtl (a); volatile long double c

Re: [PATCH] doc: update timer_* platform portability notes

Pádraig Brady wrote: > +2012-03-13 Pádraig Brady > + > + doc: Update timer_* platform portability notes. > + * doc/posix-functions/timer_create.texi: Add platforms (OpenBSD 4.9) > + that always return ENOSYS. > + * doc/posix-functions/timer_delete.texi: Likewise. > + * doc/po

[PATCH] doc: update timer_* platform portability notes

* doc/posix-functions/timer_create.texi: Add platforms that always return ENOSYS (OpenBSD 4.9). * doc/posix-functions/timer_delete.texi: Likewise. * doc/posix-functions/timer_gettime.texi: Likewise. * doc/posix-functions/timer_settime.texi: Likewise. --- ChangeLog |

Re: new module 'cbrtl'

> 2012-02-29 Bruno Haible > > New module 'cbrtl'. > * lib/math.in.h (cbrtl): New declaration. > * lib/cbrtl.c: New file. > * m4/cbrtl.m4: New file. > * m4/math_h.m4 (gl_MATH_H): Test whether cbrtl is declared. > (gl_MATH_H_DEFAULTS): Initialize GNULIB_CBRTL,

Re: new module 'remainderl'

> 2012-02-25 Bruno Haible > > New module 'remainderl'. > * lib/math.in.h (remainderh): New declaration. > * lib/remainderl.c: New file. > * m4/remainderl.m4: New file. > * modules/remainderl: New file. > * m4/math_h.m4 (gl_MATH_H): Test whether remainderl is

Re: MSVC fails to compile floor.c with -fp:strict

Michael Goffioul wrote: > Same problem occurs in cbrt.c and cbrtf.c. Fixed. No other occurrences of this problem found so far. 2012-03-13 Bruno Haible Avoid compilation errors with MSVC option -fp:strict. * lib/cbrt.c: Use MSVC specific pragma fenv_access. * lib/cbrtf

Re: [PATCH] quote: fuse into quotearg

On 03/13/2012 11:58 AM, Akim Demaille wrote: > Could you please install the following patch? Thanks, I pushed that into gnulib. The patch was corrupted, though: initial spaces were removed from lines, which made the patch format invalid, which meant that 'git am' didn't work for me and I had to

Re: [PATCH] quote: fuse into quotearg

Hi Paul, Could you please install the following patch? Le 9 mars 2012 à 08:55, Akim Demaille a écrit : > Hi Bruno, > > Le 7 mars 2012 à 11:40, Bruno Haible a écrit : > >> 2012-03-07 Bruno Haible >> >> quote: Adhere to common module description layout. >> * modules/quote (Makefile

Re: should an empty "pkgdata_DATA" cause creation of $(pkgdatadir) by "make install"?

On 03/13/12 07:30, Stefano Lattarini wrote: [CC:ing Ralf, as I'd like to hear his opinion here] Reference: On 03/13/2012 01:14 PM, Stefano Lattarini wrote: Now that I think about it, I'm not sure whether it was done "by de

Re: should an empty "pkgdata_DATA" cause creation of $(pkgdatadir) by "make install"?

On 03/13/2012 08:55 AM, Robert Boehne wrote: >> >> So we're in a sort of a tie here: some users think that the current >> Automake >> behaviour is a feature (and I lean toward that position), other ones >> (with >> Ralf among them, apparently) believe it's a bug. Hmmm. What now? >> >> Regards, >>

should an empty "pkgdata_DATA" cause creation of $(pkgdatadir) by "make install"? (was: Re: [PATCH] gnulib-tool: fix imprecise comments)

[CC:ing Ralf, as I'd like to hear his opinion here] Reference: On 03/13/2012 01:14 PM, Stefano Lattarini wrote: > > Now that I think about it, I'm not sure whether it was done "by design" from > the beginning, > And I also miss

Re: [PATCH] gnulib-tool: fix imprecise comments

Hi Bruno. On 03/13/2012 11:51 AM, Bruno Haible wrote: > Stefano Lattarini wrote: >> * gnulib-tool (func_emit_lib_Makefile_am): It's not just Automake >> versions < 1.9b that creates an empty pkgdatadir at installation >> time if pkgdata_DATA is specified to empty; modern automake versions >> do th

Re: [PATCH] gnulib-tool: fix imprecise comments

Stefano Lattarini wrote: > * gnulib-tool (func_emit_lib_Makefile_am): It's not just Automake > versions < 1.9b that creates an empty pkgdatadir at installation > time if pkgdata_DATA is specified to empty; modern automake versions > do this as well, and that's considered a feature, not a bug. Adju

[PATCH] gnulib-tool: fix imprecise comments

* gnulib-tool (func_emit_lib_Makefile_am): It's not just Automake versions < 1.9b that creates an empty pkgdatadir at installation time if pkgdata_DATA is specified to empty; modern automake versions do this as well, and that's considered a feature, not a bug. Adjust a comment accordingly. (func_e