Re: [Bug 328688] Re: [dapper] socket.c:1616: INSIST(!sock->pending_recv) failed [fixed 9.3.5-P2]

Hi Sorry I ment to update. In testing 9.3.5-P2 doesnt fix this issue. I ended up having to goto 9.3.6-P1. jim LaMont Jones wrote [03.18.09]: > ** Summary changed: > > - [dapper] socket.c:1616: INSIST(!sock->pending_recv) failed > + [dapper] socket.c:1616: INSIST(!sock->pending_recv) failed

[Bug 328688] [NEW] [dapper] socket.c:1616: INSIST(!sock->pending_recv) failed

Public bug reported: Binary package hint: bind9 Hi Ever since the the 1:9.3.2-2ubuntu1.5 update to fix the source port randomization issue we have been seeing the following error on some of our busier name servers. named[765]: socket.c:1616: INSIST(!sock->pending_recv) failed named[765]: exitin

[Bug 1588915] [NEW] network dependent services starting before network is up

Public bug reported: Description:Ubuntu 16.04 LTS (server) Release:16.04 systemd: Installed: 229-4ubuntu6 Candidate: 229-4ubuntu6 Version table: *** 229-4ubuntu6 500 500 http://us.archive.ubuntu.com/ubuntu xenial-updates/main amd64 Packages 100 /var/lib/dpkg/st

[Bug 1588915] Re: network dependent services starting before network is up

Hi openntpd was just one example of the issue. I tried setting After =network-online.target before opening this bug, it has no effect and is why I mentioned "xenial server (dunno about desktop) doesn't seem to use network-online.target". ubuntu@xenial64:~$ systemctl list-dependencies network-onl

[Bug 1588915] Re: network dependent services starting before network is up

Hi The logged output is typical of tg3 a physical interface. The first time you bring the interface up the kernel will log that the link is down, which it is. boot server with no interfaces configured root@xenial64:~# ifconfig loLink encap:Local Loopback inet addr:127.0.0.1 M

[Bug 1588915] Re: networking.service does not wait for interfaces to be up

As I mention in the beginning the issue is likely specific to a physical box configured with a static ip. If a box is configured for dhcp, networking.service cant complete the configuration of the interface until its gotten an dhcp address. The act of getting dhcp address would require the interf

[Bug 956051] Re: libc6 crash while running 'xm'

HI The attached patch against 2.15-0ubuntu10 resolves the issue for me. It does this by disabling FMA4 if AVX is unavailable. thanks jim ** Patch added: "fma4-depends-avx.diff" https://bugs.launchpad.net/ubuntu/+source/eglibc/+bug/956051/+attachment/3135508/+files/fma4-depends-avx.diff --

[Bug 956051] Re: libc6 crash while running 'xm'

Hi It appears this issue can be worked around by adding the following options to xen's command line in grub. cpuid_mask_ext_ecx=0xFFFEFFF cpuid_mask_ecx=0xEFF This masks both FMA4 and AVX from cpuid for dom0 and all domU's Also note that xen has a cpuid option for domU configs that allows t

[Bug 956051] Re: libc6 crash while running 'xm'

Hi In briefly talking with the xen folks, this is a glibc bug. I filed the following bug report with them. http://sourceware.org/bugzilla/show_bug.cgi?id=14059 jim ** Bug watch added: Sourceware.org Bugzilla #14059 http://sourceware.org/bugzilla/show_bug.cgi?id=14059 -- You received this

[Bug 956051] Re: libc6 crash while running 'xm'

I am running into this issue as well on a 12.04 domU with a AMD Opteron 6272 based box. [14553.237586] apport[18522] trap invalid opcode ip:7f1d632625fc sp:78701390 error:0 in libm-2.15.so[7f1d6322+f9000] Debugging info (gdb) exec-file python-dbg (gdb) run Starting program: /usr/bin/pyth

[Bug 956051] Re: libc6 crash while running 'xm'

Hi Some additional information that leads me to believe this is a xen issue. gcc 4.6.3 code has the following #define OPTION_MASK_ISA_FMA4_SET \ (OPTION_MASK_ISA_FMA4 | OPTION_MASK_ISA_SSE4A_SET \ | OPTION_MASK_ISA_AVX_SET) So when you use -mfma4 option it also enables avx instruction set.