On Tue, Apr 01, 2008 at 06:32:36PM +0200, Gerfried Fuchs wrote: > On Mon, Mar 31, 2008 at 02:29:06PM -0600, dann frazier wrote: > > Our automated buildd log filter[1] detected a problem that is likely to > ^^^^^^ > > cause your package to segfault on architectures where the size of a > > pointer is greater than the size of an integer, such as ia64 and amd64. > > Did you confirm that it's really like that in any way?
If a pointer returned by one of those functions is dereferenced, it will segfault on ia64. If you have a simple way for me to setup such a scenario, I can do so. Ignoring a returned pointer from a function is a corner case though, so please understand that since I file so many of these issues that I don't attempt to prove them in every case. http://bugs.debian.org/cgi-bin/[EMAIL PROTECTED];which=tag&data=implicit-pointer-conversion&exclude=sarge&archive=no > > Function `ldap_init' implicitly converted to pointer at etpan-ldap.c:127 > > Function `ldap_get_values' implicitly converted to pointer at > > etpan-ldap.c:346 > > Alright, implicit conversions are often a bad thing anyway. > > On Mon, Mar 31, 2008 at 03:17:49PM -0600, dann frazier wrote: > > Since this is an RC bug, I've gone ahead and uploaded an NMU. This > > qualifies as a 0-day NMU, but since this bug has just been submitted > > and the maintainer is not on the LowThresholdNMU list, I've elected to > > upload to DELAYED/7-day instead. Patch is attached. > > I am slightly offended by the rush of a "likely" problem here, doing it > within not even an hour, especially since it was also you who reported > the problem but originally calling it a likely problem which means one > should rather check if it's a real one. It is *absolutely* a problem that this package will FTBFS on ia64 due to a buildd hook to fail such builds. This is the reason for the seriousness and the NMU (which is in DELAYED/7-day, not in incoming). > This feels much more like anything but a hijack. Feel free to upload > with setting yourself as Maintainer too if you feel so strongly about > it. I thought it polite to do a delayed upload instead of a 0-day NMU, sorry you feel differently. Please don't take it as an attack, hijack, etc - I just preferred to do it this way than to wait a week and do a 0-day NMU because it saves time for me to do it all at once, and the end result is the same: you can wait a week for it to be automatically fixed, or upload an MU instead. -- dann frazier -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
