-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
According to jakub at redhat dot com on 12/21/2007 7:00 AM:
> --- Additional Comments From jakub at redhat dot com 2007-12-21 14:00
> ---
>> Can you explain why?
>
> Because the malloc call makes the function no longer async-signal safe.
> W
Le mercredi 2 janvier 2008, Bruno Haible a écrit :
> Hello,
>
> Bastien Roucaries wrote:
> > I need for my program to use asinh acosh and atanh. For now I used to
> > define
> >
> > it if missing using
> > #define asinh(x) log(x+sqrt(x*x+1))
> > #define acosh(x) log(x+sqrt(x-1.0)*sqrt(x+1.0))
> > #
Hello Colin,
Colin Watson wrote:
> I decided to write some documentation for this (diff attached), which I
> hope you can review and add to Gnulib.
Many thanks for this! I changed the node title and a few other details and
committed
the change as appended.
> If you require
> a copyright assignm
Paul Eggert wrote:
> Yes, thanks, that looks good.
Committed.
Bruno
"Bruno Haible" <[EMAIL PROTECTED]> writes:
> * lib/memmem.c (knuth_morris_pratt, memmem): Change all 'char *'
> variables to 'unsigned char *' type.
> Reported by Paul Eggert.
Yes, thanks, that looks good.
> Thanks, but the email that I got lacked a patch.
Sorry about it. Here it is (please forgive the line breaking caused by the
webmailer):
2008-01-02 Bruno Haible <[EMAIL PROTECTED]>
* lib/memmem.c (knuth_morris_pratt, memmem): Change all 'char *'
variables to 'unsigned char *
"Bruno Haible" <[EMAIL PROTECTED]> wrote:
> Paul Eggert wrote:
>> section 6.2.6.1:
>> ...
>> Here again, the intent is that memcmp(&x, &y, sizeof (T)) compares all
>> the bits in the underlying representation. The only way to do that
>> portably is via "unsigned char *".
>
> OK. Here's a proposed
