Re: [Qemu-devel] Race condition in char device setup causing SEGV

2012-08-23 Thread Anthony Liguori
"Daniel P. Berrange" writes: > When testing with the new "-M none" arg, I've noticed that ~70% > of the time libvirt starts QEMU will result in a SEGV from QEMU > with the following stack trace: > > (gdb) bt > #0 0x in ?? () > #1 0x5567a37f in json_lexer_feed_char (lexer

Re: [Qemu-devel] Race condition in char device setup causing SEGV

2012-08-23 Thread Daniel P. Berrange
On Thu, Aug 23, 2012 at 12:01:05PM +0100, Daniel P. Berrange wrote: > When testing with the new "-M none" arg, I've noticed that ~70% > of the time libvirt starts QEMU will result in a SEGV from QEMU > with the following stack trace: > > (gdb) bt > #0 0x in ?? () > #1 0x5

[Qemu-devel] Race condition in char device setup causing SEGV

2012-08-23 Thread Daniel P. Berrange
When testing with the new "-M none" arg, I've noticed that ~70% of the time libvirt starts QEMU will result in a SEGV from QEMU with the following stack trace: (gdb) bt #0 0x in ?? () #1 0x5567a37f in json_lexer_feed_char (lexer=0x5658fb20, ch=123 '{', flush=false) a