Bug#351629: Still in 2.3.6.ds1-10 with sarge binutils
On Sun, Mar 04, 2007 at 08:22:10PM -0600, Michael Koren wrote: Hi, I don't know if I can reopen this as a normal user, but I can still reproduce this with libc6-dev from etch and binutils from sarge, that's an unsupported mix. there is a gcc-3.3 in etch, binutils in etch, please install this combination. I was referring to the problem described by the orignal reporter, i.e., which gcc-3.3 doesn't matter, but post-sarge binutils doesn't work: Error also goes away if binutils is upgraded from sarge version to version currently in unstable. But this upgrade in turn causes problems with usage of g++ 3.3, described at http://gcc.gnu.org/bugzilla/show_bug.cgi?id=16625. This problem happens on Debian if binutils are post-sarge, with any version of g++-3.3 packages. So at least, dependency on recent binutils should be added. But it would be much better if compatibility with sarge binutils will be preserved in libc6-dev until a workable combination of g++-3.3 and post-sarge binutils will be found. Unfortunately, we do have to use g++-3.3 ABI here, and I'm sure we are not alone with that. I guess this is really a bug in binutils, but it still means this package is apparently incompatible with g++-3.3 for now, due to a few mysterious name changes in libc_nonshared.a. That leaves no suitable package in etch, hence my two questions above. Thanks, Michael P.S. Should a copy of this bug be sent to the binutils package? I think so. Though I thought that gcc-3.3 would be removed in etch, I think it's only here for java :| -- ·O· Pierre Habouzit ··O[EMAIL PROTECTED] OOOhttp://www.madism.org pgpMuZaekHqcX.pgp Description: PGP signature
r2020 - in glibc-package/trunk/debian: . patches patches/any
Author: madcoder Date: 2007-03-05 09:34:52 + (Mon, 05 Mar 2007) New Revision: 2020 Added: glibc-package/trunk/debian/patches/any/cvs-sunrpc-64bits-svc_getreqset.diff Modified: glibc-package/trunk/debian/changelog glibc-package/trunk/debian/patches/series Log: Closes: 413450 * any/cvs-sunrpc-64bits-svc_getreqset.diff: fix 64bits issue for 64bits machines (affects mount, causes CPU hog if triggered). Closes: #413450. Signed-off-by: Pierre Habouzit [EMAIL PROTECTED] Modified: glibc-package/trunk/debian/changelog === --- glibc-package/trunk/debian/changelog2007-03-04 00:00:58 UTC (rev 2019) +++ glibc-package/trunk/debian/changelog2007-03-05 09:34:52 UTC (rev 2020) @@ -1,3 +1,11 @@ +glibc (2.3.6.ds1-14) UNRELEASED; urgency=low + + [ Pierre Habouzit ] + * any/cvs-sunrpc-64bits-svc_getreqset.diff: fix 64bits issue for 64bits +machines (affects mount, causes CPU hog if triggered). Closes: #413450. + + -- Pierre Habouzit [EMAIL PROTECTED] Mon, 5 Mar 2007 10:32:34 +0100 + glibc (2.3.6.ds1-13) unstable; urgency=low * debhelper.in/libc.postinst: try to add /etc/ld.so.conf.d support for Added: glibc-package/trunk/debian/patches/any/cvs-sunrpc-64bits-svc_getreqset.diff === --- glibc-package/trunk/debian/patches/any/cvs-sunrpc-64bits-svc_getreqset.diff 2007-03-04 00:00:58 UTC (rev 2019) +++ glibc-package/trunk/debian/patches/any/cvs-sunrpc-64bits-svc_getreqset.diff 2007-03-05 09:34:52 UTC (rev 2020) @@ -0,0 +1,13 @@ +Index: sunrpc/svc.c +=== +--- sunrpc/svc.c.orig sunrpc/svc.c +@@ -372,7 +372,7 @@ + setsize = FD_SETSIZE; + maskp = readfds-fds_bits; + for (sock = 0; sock setsize; sock += NFDBITS) +-for (mask = *maskp++; (bit = ffs (mask)); mask ^= (1 (bit - 1))) ++for (mask = *maskp++; (bit = ffsl (mask)); mask ^= (1 (bit - 1))) + INTUSE(svc_getreq_common) (sock + bit - 1); + } + INTDEF (svc_getreqset) Modified: glibc-package/trunk/debian/patches/series === --- glibc-package/trunk/debian/patches/series 2007-03-04 00:00:58 UTC (rev 2019) +++ glibc-package/trunk/debian/patches/series 2007-03-05 09:34:52 UTC (rev 2020) @@ -186,3 +186,4 @@ any/submitted-strfmon.diff any/cvs-sunrpc-xdrmem_setpos.diff any/cvs-getent-wrong-struct-size.diff -p0 +any/cvs-sunrpc-64bits-svc_getreqset.diff -p0 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
r2021 - in glibc-package/trunk/debian: . control.in rules.d
Author: madcoder
Date: 2007-03-05 09:45:00 + (Mon, 05 Mar 2007)
New Revision: 2021
Modified:
glibc-package/trunk/debian/changelog
glibc-package/trunk/debian/control
glibc-package/trunk/debian/control.in/main
glibc-package/trunk/debian/rules.d/debhelper.mk
Log:
Fixes the nscd dependencies problem once for all.
* control.in/main, rules.d/debhelper.mk: Backport fix from Aur?\195?\169lien
in the
glibc-2.5 branch: use dh_shlibdeps to set the dependencies of nscd.
Closes: #409288, #412562.
Signed-off-by: Pierre Habouzit [EMAIL PROTECTED]
Modified: glibc-package/trunk/debian/changelog
===
--- glibc-package/trunk/debian/changelog2007-03-05 09:34:52 UTC (rev
2020)
+++ glibc-package/trunk/debian/changelog2007-03-05 09:45:00 UTC (rev
2021)
@@ -3,6 +3,9 @@
[ Pierre Habouzit ]
* any/cvs-sunrpc-64bits-svc_getreqset.diff: fix 64bits issue for 64bits
machines (affects mount, causes CPU hog if triggered). Closes: #413450.
+ * control.in/main, rules.d/debhelper.mk: Backport fix from Aurélien in the
+glibc-2.5 branch: use dh_shlibdeps to set the dependencies of nscd.
+Closes: #409288, #412562.
-- Pierre Habouzit [EMAIL PROTECTED] Mon, 5 Mar 2007 10:32:34 +0100
Modified: glibc-package/trunk/debian/control
===
--- glibc-package/trunk/debian/control 2007-03-05 09:34:52 UTC (rev 2020)
+++ glibc-package/trunk/debian/control 2007-03-05 09:45:00 UTC (rev 2021)
@@ -58,7 +58,7 @@
Architecture: alpha amd64 arm arm-softfloat armeb i386 m32r m68k mips mipsel
powerpc ppc64 sparc ia64 hppa s390 sh3 sh4 sh3eb sh4eb kfreebsd-i386
kfreebsd-amd64
Section: admin
Priority: optional
-Depends: libc6 (= ${Source-Version})
+Depends: ${shlibs:Depends}
Description: GNU C Library: Name Service Cache Daemon
A daemon which handles passwd, group and host lookups
for running programs and caches the results for the next
Modified: glibc-package/trunk/debian/control.in/main
===
--- glibc-package/trunk/debian/control.in/main 2007-03-05 09:34:52 UTC (rev
2020)
+++ glibc-package/trunk/debian/control.in/main 2007-03-05 09:45:00 UTC (rev
2021)
@@ -58,7 +58,7 @@
Architecture: @threads_archs@
Section: admin
Priority: optional
-Depends: @libc@ (= ${Source-Version})
+Depends: ${shlibs:Depends}
Description: GNU C Library: Name Service Cache Daemon
A daemon which handles passwd, group and host lookups
for running programs and caches the results for the next
Modified: glibc-package/trunk/debian/rules.d/debhelper.mk
===
--- glibc-package/trunk/debian/rules.d/debhelper.mk 2007-03-05 09:34:52 UTC
(rev 2020)
+++ glibc-package/trunk/debian/rules.d/debhelper.mk 2007-03-05 09:45:00 UTC
(rev 2021)
@@ -125,7 +125,9 @@
dh_makeshlibs -p$(curpass) -V $(call xx,shlib_dep)
dh_installdeb -p$(curpass)
- # dh_shlibdeps -p$(curpass)
+ if [ $(curpass) = nscd ] ; then \
+ dh_shlibdeps -p$(curpass) ; \
+ fi
dh_gencontrol -p$(curpass) -- $($(curpass)_control_flags)
dh_md5sums -p$(curpass)
dh_builddeb -p$(curpass)
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Processed: sys/inotify.h
Processing commands for [EMAIL PROTECTED]: reopen 369402 Bug#369402: libc6-dev: inotify-syscalls.h is missing Bug reopened, originator not changed. thanks Stopping processing here. Please contact me if you need assistance. Debian bug tracking system administrator (administrator, Debian Bugs database) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#369402: sys/inotify.h
reopen 369402 thanks There is quite a lot software out there today which uses inotify. Each of this package has to be updated to ship a private copy of the inotify syscall. Using linux/inotify.h is not a good option. Please consider to patch/backport sys/inotify from current upstream glibc.h, so this software can be compiled unaltered. Cheers, Michael -- Why is it that all of the instruments seeking intelligent life in the universe are pointed away from Earth? signature.asc Description: OpenPGP digital signature
Processed: severity of 369402 is wishlist
Processing commands for [EMAIL PROTECTED]: # Automatically generated email from bts, devscripts version 2.9.27 severity 369402 wishlist Bug#369402: libc6-dev: inotify-syscalls.h is missing Severity set to `wishlist' from `normal' End of message, stopping processing here. Please contact me if you need assistance. Debian bug tracking system administrator (administrator, Debian Bugs database) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#369402: marked as done (libc6-dev: inotify-syscalls.h is missing)
Your message dated Mon, 05 Mar 2007 19:35:55 +0100 with message-id [EMAIL PROTECTED] and subject line Bug#369402: sys/inotify.h has caused the attached Bug report to be marked as done. This means that you claim that the problem has been dealt with. If this is not the case it is now your responsibility to reopen the Bug report if necessary, and/or fix the problem forthwith. (NB: If you are a system administrator and have no idea what I am talking about this indicates a serious mail system misconfiguration somewhere. Please contact me immediately.) Debian bug tracking system administrator (administrator, Debian Bugs database) ---BeginMessage--- Package: libc6-dev Version: 2.3.6-10 Severity: normal The file inotify-syscalls.h from http://www.kernel.org/pub/linux/kernel/people/rml/inotify/headers/ is missing. This prevents its usage in applications, except for those that right now have the habit to ship private copies, which is bad. See https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=179366 for a similar case. Either the file should be shipped with libc6-dev, or (I think this was done upstream, not sure, but preferrable) the syscalls should be merged into the bits/syscalls.h file. -- System Information: Debian Release: testing/unstable APT prefers unstable APT policy: (500, 'unstable') Architecture: powerpc (ppc) Kernel: Linux 2.6.14-2-powerpc Locale: LANG=pt_BR.UTF-8, LC_CTYPE=pt_BR.UTF-8 (charmap=UTF-8) Versions of packages libc6-dev depends on: ii libc62.3.6-10GNU C Library: Shared libraries ii linux-kernel-headers 2.6.13+0rc3-1.1 Linux Kernel Headers for developme -- no debconf information ---End Message--- ---BeginMessage--- Version: 2.5-0exp6 Michael Biebl a écrit : reopen 369402 thanks There is quite a lot software out there today which uses inotify. Each of this package has to be updated to ship a private copy of the inotify syscall. Using linux/inotify.h is not a good option. Please consider to patch/backport sys/inotify from current upstream glibc.h, so this software can be compiled unaltered. Etch is frozen nothing can be added. Please use glibc 2.5 from experimental. Closing the bug with the version from experimental. Cheers, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net ---End Message---
Bug#413450: libc6: svc_getreqset doesn't work on 64bit hosts
Version: 2.5-0exp6 Neil Brown a écrit : Package: libc6 Version: 2.3.6.ds1-13 Severity: normal Tags: patch On 64bit machines, svc_getreqset ignores filedescriptors 32-63 and others. This has been fixed in glibc 2.4 but needs to be fixed in 2.3.6.debian while we keep using it. The consequence of this bug is that if mountd (or any other rpc server) gets 26 or more concurrent tcp connections, it goes into a spin constantly using CPU, and not servicing requests on those connections. Following patch fixes it. Thanks, the patch has been added to our SVN, but it probably won't be fixed until the release of Debian 4.0r1. I am closing the bug with the version from experimental, which has the fix. -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#369402: sys/inotify.h
Aurelien Jarno wrote: Version: 2.5-0exp6 Michael Biebl a écrit : reopen 369402 thanks There is quite a lot software out there today which uses inotify. Each of this package has to be updated to ship a private copy of the inotify syscall. Using linux/inotify.h is not a good option. Please consider to patch/backport sys/inotify from current upstream glibc.h, so this software can be compiled unaltered. Etch is frozen nothing can be added. Please use glibc 2.5 from experimental. Closing the bug with the version from experimental. Well then, let's hope we get out etch soon and glibc 2.5 in unstable. I wanted to start working on hal-0.5.9 which has the above problem and I wanted to avoid having to patch hal for that. upstream has no interest to support glibc 2.4. I already did this dance with tracker and upstart and don't want to do it again. I guess hal will have to wait then... Cheers, Michael -- Why is it that all of the instruments seeking intelligent life in the universe are pointed away from Earth? signature.asc Description: OpenPGP digital signature
Bug#413450: marked as done (libc6: svc_getreqset doesn't work on 64bit hosts)
Your message dated Mon, 05 Mar 2007 19:40:05 +0100 with message-id [EMAIL PROTECTED] and subject line Bug#413450: libc6: svc_getreqset doesn't work on 64bit hosts has caused the attached Bug report to be marked as done. This means that you claim that the problem has been dealt with. If this is not the case it is now your responsibility to reopen the Bug report if necessary, and/or fix the problem forthwith. (NB: If you are a system administrator and have no idea what I am talking about this indicates a serious mail system misconfiguration somewhere. Please contact me immediately.) Debian bug tracking system administrator (administrator, Debian Bugs database) ---BeginMessage--- Package: libc6 Version: 2.3.6.ds1-13 Severity: normal Tags: patch On 64bit machines, svc_getreqset ignores filedescriptors 32-63 and others. This has been fixed in glibc 2.4 but needs to be fixed in 2.3.6.debian while we keep using it. The consequence of this bug is that if mountd (or any other rpc server) gets 26 or more concurrent tcp connections, it goes into a spin constantly using CPU, and not servicing requests on those connections. Following patch fixes it. Thanks, NeilBrown ### Diffstat output ./sunrpc/svc.c |2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff .prev/sunrpc/svc.c ./sunrpc/svc.c --- .prev/sunrpc/svc.c 2007-03-05 14:50:06.0 +1100 +++ ./sunrpc/svc.c 2007-03-05 14:50:44.0 +1100 @@ -372,7 +372,7 @@ svc_getreqset (fd_set *readfds) setsize = FD_SETSIZE; maskp = readfds-fds_bits; for (sock = 0; sock setsize; sock += NFDBITS) -for (mask = *maskp++; (bit = ffs (mask)); mask ^= (1 (bit - 1))) +for (mask = *maskp++; (bit = ffsl (mask)); mask ^= (1L (bit - 1))) INTUSE(svc_getreq_common) (sock + bit - 1); } INTDEF (svc_getreqset) -- System Information: Debian Release: 4.0 APT prefers unstable APT policy: (500, 'unstable') Architecture: amd64 (x86_64) Shell: /bin/sh linked to /bin/bash Kernel: Linux 2.6.18-4-amd64 Locale: LANG=en_AU.UTF-8, LC_CTYPE=en_AU.UTF-8 (charmap=UTF-8) Versions of packages libc6 depends on: ii tzdata2007b-1Time Zone and Daylight Saving Time libc6 recommends no packages. -- no debconf information ---End Message--- ---BeginMessage--- Version: 2.5-0exp6 Neil Brown a écrit : Package: libc6 Version: 2.3.6.ds1-13 Severity: normal Tags: patch On 64bit machines, svc_getreqset ignores filedescriptors 32-63 and others. This has been fixed in glibc 2.4 but needs to be fixed in 2.3.6.debian while we keep using it. The consequence of this bug is that if mountd (or any other rpc server) gets 26 or more concurrent tcp connections, it goes into a spin constantly using CPU, and not servicing requests on those connections. Following patch fixes it. Thanks, the patch has been added to our SVN, but it probably won't be fixed until the release of Debian 4.0r1. I am closing the bug with the version from experimental, which has the fix. -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net ---End Message---
r2022 - glibc-package/branches/glibc-2.5/debian
Author: aurel32 Date: 2007-03-05 19:13:15 + (Mon, 05 Mar 2007) New Revision: 2022 Modified: glibc-package/branches/glibc-2.5/debian/changelog Log: - Provides sys/inotify.h. Closes: #369402. Modified: glibc-package/branches/glibc-2.5/debian/changelog === --- glibc-package/branches/glibc-2.5/debian/changelog 2007-03-05 09:45:00 UTC (rev 2021) +++ glibc-package/branches/glibc-2.5/debian/changelog 2007-03-05 19:13:15 UTC (rev 2022) @@ -8,6 +8,7 @@ - Fixes strtod wrt hex floats with negative exponent. Closes: #172562. - Fixes nice() errno when called without appropriate privileges. Closes: #286825. +- Provides sys/inotify.h. Closes: #369402. [ Aurelien Jarno ] * debian/shlibver: Bump up to 2.5. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#408850: mksh: FTBFS on experimental/alpha (Re: Log for failed build of mksh_28.9.20070118 (dist=experimental))
I have an idea…
Dixi:
Martin Zobel-Helas dixit:
/usr/include/sys/stat.h:217: error: conflicting types for 'stat'
/usr/include/sys/stat.h:365: error: previous definition of 'stat' was here
[...]
Sounds like a bug in the header file to me:
215 extern int __REDIRECT_NTH (stat, (__const char *__restrict __file,
216 struct stat *__restrict __buf),
stat64)
217 __nonnull ((1, 2));
363 extern __inline__ int
364 __NTH (stat (__const char *__path, struct stat *__statbuf))
365 {
366 return __xstat (_STAT_VER, __path, __statbuf);
367 }
The second one is missing the restrict qualifier. Could that
be the cause, possibly induced by invoking the C99 standard
during compiling with -std=gnu99 ?
Could please someone with access to an alpha try to build the version
of mksh currently in experimental, or even my “current playchild“
at http://users.unixforge.de/~tglaser/mksh_28.9.20070304.dsc, with
HAVE_CAN_STDG99=0 defined in the environment (of Build.sh, i.e. I'd
patch debian/rules appropriately) and show me the build log?
If so, we might at least have a workaround until that's fixed in
the system headers, because I've got a major update to mksh (i.e.
a new formal release) I'd like to get into unstable/testing, and
(thanks to playing around in experimental) only this bug is pre-
venting me from doing so.
@Pierre HABOUZIT: maybe your sample programme compiles fine as
long as you don't use -std=gnu99? Maybe the issue only appears
on alpha because it implies some kind of alignment changes?
Thanks!
bye,
//mirabile
--
I believe no one can invent an algorithm. One just happens to hit upon it
when God enlightens him. Or only God invents algorithms, we merely copy them.
If you don't believe in God, just consider God as Nature if you won't deny
existence. -- Coywolf Qi Hunt
