Date: Wed, 09 Mar 2016 08:39:58 +0100 From: Nils Beyer <nbe@renzel.net> To: freebsd-emulation@freebsd.org Subject: Re: citrix_ica - "wfica" generates deadlock in linux_sys_futex... Message-ID: <5463752.mgVZH8y5u3@asbach.renzel.net> References: <1710849.HCaPmFIAXo@asbach.renzel.net> <20160308075527.GA29364@chd.heemeyer.club> <1794078.TnZCHkfeGo@asbach.renzel.net> <20160308085140.GA30833@chd.heemeyer.club> <2225619.DhA8sIh2h8@asbach.renzel.net> <20160308093009.GA31396@chd.heemeyer.club> <20160308115018.GM67250@kib.kiev.ua>
next in thread | previous in thread | raw e-mail | index | archive | help
Hi Konstantin, Konstantin Belousov wrote: >> this is strange as no ipc changes are made in stable/10 > > There was one interesting change, AFAIR. > > Check the value of sysctl kern.ipc.shm_allow_removed, and if it's value > is one, change it to zero. unfortunately, that doesn't make any difference. Nevertheless, I need that sysctl enabled for "chromium". And it needs to be enabled for "wfica" under 10.2 kernel as well. For what it's worth, here's "procstat -kk <PID of wfica>": ============================================================================== PID TID COMM TDNAME KSTACK 1760 100731 wfica - mi_switch+0xe1 sleepq_catch_signals+0xab sleepq_wait_sig+0xf _sleep+0x27d linux_sys_futex+0xb3c ia32_syscall+0x3a0 Xint0x80_syscall+0x95 1760 100732 wfica - mi_switch+0xe1 sleepq_catch_signals+0xab sleepq_wait_sig+0xf _sleep+0x27d sys_semop+0x53f linux_semop+0x25 linux_ipc+0x4b ia32_syscall+0x3a0 Xint0x80_syscall+0x95 ============================================================================== Regads, Nils
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?5463752.mgVZH8y5u3>