Re: changing "configure" to default to "gcc -g -O2 -fwrapv ..."

Ian Lance Taylor wrote: The new option -fstrict-overflow tells gcc that it can assume the strict signed overflow semantics prescribed by the language standard. This option is enabled by default at -O2 and higher. Using -fno-strict-overflow will tell gcc that it can not assume that signed overfl

Re: changing "configure" to default to "gcc -g -O2 -fwrapv ..."

Paul Eggert <[EMAIL PROTECTED]> writes: > >> GCC itself relies on wrapv semantics. As does glibc. And > >> coreutils. And GNU tar. And Python. I'm sure there are > >> many other significant programs. I don't have time to do a > >> comprehensive survey right now. > > > > Where does GCC rely o

Re: adding bootstrap and bootstrap.conf

Eric Blake wrote: > > Find attached a patch that implements the first part of it. The > > second part will be the handling of CLEANFILES and MOSTLYCLEANFILES. > > No attachment, but I see you applied the patch? Indeed, the attachment got lost when I had to resend the mail, due to earlier mail pro

Re: source(builtin) and read(2)

Eric Blake wrote: > I'm seeking backup from those more knowledgeable about the C > and POSIX standards as to whether NSK is allowed to define SSIZE_MAX to > something smaller than what the underlying type can hold I'm less knowledgeable than Paul, but I would say that 52*1024 is not an "appropria

Re: new module write-any-file

Hello Paul, Please accept some nitpicking: - The code uses the Solaris 10 APIs. How about also using the Linux API for "capabilities" - capget() - that was added in Linux 2.2 and glibc-2.1.x? - If priv_allocset fails, the function returns false without considering geteuid (). And if priv_alloc

Re: source(builtin) and read(2)

Eric Blake <[EMAIL PROTECTED]> writes: > I'm seeking backup from those more knowledgeable about the C > and POSIX standards as to whether NSK is allowed to define SSIZE_MAX to > something smaller than what the underlying type can hold, Yes, that's my understanding. To back this up,

Re: source(builtin) and read(2)

-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 [originally mentioned on bug-bash] According to Matthew Woehlke on 3/23/2007 3:40 PM: > Eric Blake wrote: >> According to Matthew Woehlke on 3/23/2007 2:40 PM: SSIZE_MAX is guaranteed to be the maximum value that fits in ssize_t. > > ...that "fi

new module write-any-file

I installed this; it should be useful for coreutils and perhaps for other applications. 2007-03-23 Paul Eggert <[EMAIL PROTECTED]> * MODULES.html.sh (File system functions): New module write-any-file. * modules/write-any-file, lib/write-any-file.c, lib/write-any-file.h:

Re: strptime.c: ANSI C compatibility

Bruno Haible <[EMAIL PROTECTED]> writes: > Here is a proposed fix. OK, Paul? Yes, please install.

avoid space-tab

-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 I'm applying this patch, because in the past, I have inadvertently corrupted files when emacs whitespace mode decided to be 'helpful' and collapse space-tab into plain tab. Thus, when writing portable sed expressions, a good rule of thumb is to use ta

Re: adding bootstrap and bootstrap.conf

-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 According to Bruno Haible on 3/19/2007 6:19 PM: > Jim Meyering wrote: >> Another feature: bootstrap adds .cvsignore and .gitignore >> entries for files it supplies. > > Good idea. It has annoyed me as well, but never reached the necessary pain > level

Re: --version output change proposal

-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 According to Karl Berry on 3/22/2007 2:30 PM: > rms wanted to use a specific url after all. Thus, please use: > > License GPLv2+: GNU GPL version 2 or later > \n\ Committed, with that change. - -- Don't work too h

Re: fnmatch has exponential running time

On 3/22/07, Bruno Haible <[EMAIL PROTECTED]> wrote: fnmatch() has a worst-case complexity O(m*n) where m is the size of the pattern and n is the size of the sample string. Unfortunately glibc has chosen an implementation with exponential running time. Yes. Oddly, per some testng I did about a