On 2012-05-09 07:11, Kevin Wolf wrote: > Am 08.05.2012 21:35, schrieb Jan Kiszka: >> Hi, >> >> I hunted down a fairly subtle corruption of the VCPU thread signal mask >> in KVM mode when using the ucontext version of coroutines: >> >> coroutine_new calls getcontext, makecontext, swapcontext. Those >> functions get/set also the signal mask of the caller. Unfortunately, >> they only use the sigprocmask syscall on i386, not the rt_sigprocmask >> version. So they do not properly save/restore the blocked RT signals, >> namely our SIG_IPI - it becomes unblocke this way. > > If other coroutine backends work (sigaltstack?),
Had a brief look at signalstack. Is kill(getpid(), SIGUSR2) in coroutine_new correct? This should broadcast to the whole process, but we only block SIGUSR2 in the caller's context. Why not pthread_kill? I will give it a try nevertheless. > we could try to detect > the situation in configure and set the right default. Not sure what the > condition is, glibc + i386? So far it looks like a glibc/i386 thing, but maybe it is driven by the ABI that defines the size of jmpbuf and, thus, the ability to store RT signals as well. Then it is a generic i386 limitation. Jan -- Siemens AG, Corporate Technology, CT T DE IT 1 Corporate Competence Center Embedded Linux
