From Hugo Dias (hdias): It seems that calling the svc_listen function in 'net/atm/svc.c' twice on same socket, will create unassigned PVC/SVC entries. This entries are visible using proc filesystem. #cat /proc/net/atm/vc Address Itf ... c7f34400 Unassigned ... c7f34400 Unassigned ... c7f34400 Unassigned ... ....... The code in 'net/atm/proc.c', can't handle the unassigned entries. Kernel will freeze with infinite loop in 'proc.c' if we cat '/proc/net/atm/pvc' : net/atm/proc.c: 074 static inline int compare_family(struct sock *sk, int family) 073 { 074 return !family || (sk->sk_family == family); 075 } 091 try_again: 092 for (; sk; sk = sk_next(sk)) { 093 l -= compare_family(sk, family); <<<<<<<<< 094 if (l < 0) 095 goto out; 096 } Version-Release number of selected component (if applicable): Linux Kernel 2.6.x (Fedora Core 10 - 2.6.25-14.fc9.i686) (Fedora Core 9 - 2.6.25-14.fc9.x86_64) (Fedora Core 9 - 2.6.25-14.fc9.i686 ) How reproducible: Please see attachment Actual results: Kernel hang Expected results: Not Hang
Seems to be introduced by upstream commit 9301e32: From 9301e320e98ff19a0e48881b038d0c24ca76e6c0 Mon Sep 17 00:00:00 2001 From: Chas Williams <chas.navy.mil> Date: Wed, 28 Sep 2005 16:35:01 -0700 Subject: [PATCH] [ATM]: track and close listen sockets when sigd exits Signed-off-by: Chas Williams <chas.navy.mil> Tested the upstream kernel without commit 9301e32. It's not reproducible.
(In reply to comment #3) > on rhel4 kernel, this bug is not triggerable. Like rhel4, tested 2.4.21-57.EL on rhel3, and the bug is not triggerable. [test@rhel3-as-i386 test]$ ./atmdos Itf VPI VCI AAL RX(PCR,Class) TX(PCR,Class) [test@rhel3-as-i386 test]$ uname -a Linux rhel3-as-i386 2.4.21-57.EL #1 Wed Apr 23 01:46:01 EDT 2008 i686 i686 i386 GNU/Linux
Created attachment 325649 [details] Proposed patch for real-time kernel taskID=1598153; Tested this with the reproducer.
This is public now. http://marc.info/?l=linux-netdev&m=122841256115780&w=2 http://marc.info/?l=linux-netdev&m=122843162615569&w=2
Fixed in Fedora 10 kernel 2.6.27.8-143
Where is the Fedora 9 bug?
(In reply to comment #19) > Where is the Fedora 9 bug? Don't think there is one. Please create it if needed. Thanks Chuck.
(In reply to comment #17) > This is public now. > http://marc.info/?l=linux-netdev&m=122841256115780&w=2 > http://marc.info/?l=linux-netdev&m=122843162615569&w=2 Upstream commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=17b24b3c97498935a2ef9777370b1151dfed3f6f
kernel-2.6.27.9-159.fc10 has been submitted as an update for Fedora 10. http://admin.fedoraproject.org/updates/kernel-2.6.27.9-159.fc10
kernel-2.6.27.9-73.fc9 has been submitted as an update for Fedora 9. http://admin.fedoraproject.org/updates/kernel-2.6.27.9-73.fc9
kernel-2.6.27.9-159.fc10 has been pushed to the Fedora 10 stable repository. If problems still persist, please make note of it in this bug report.
kernel-2.6.27.9-73.fc9 has been pushed to the Fedora 9 stable repository. If problems still persist, please make note of it in this bug report.
kernel-2.6.26.8-57.fc8 has been submitted as an update for Fedora 8. http://admin.fedoraproject.org/updates/kernel-2.6.26.8-57.fc8
kernel-2.6.26.8-57.fc8 has been pushed to the Fedora 8 stable repository. If problems still persist, please make note of it in this bug report.
This was addressed via: Red Hat Enterprise Linux (v. 5.2.z server) (RHSA-2009:0021) MRG Realtime for RHEL 5 Server (RHSA-2009:0053) Red Hat Enterprise Linux version 5 (RHSA-2009:0225)