Re: [Xen-devel] [PATCH 9/9] libxl: Kill QEMU with "reaper" ruid

2018-11-29 Thread Ian Jackson
George Dunlap writes ("Re: [PATCH 9/9] libxl: Kill QEMU with "reaper" ruid"): > On Nov 28, 2018, at 5:02 PM, Ian Jackson wrote: > > Also I think you need to handle errors properly ? Ie set and check > > errno. > > Don’t I want to pass up the errno values set by the getpwnam functions? By `set'

Re: [Xen-devel] [PATCH 9/9] libxl: Kill QEMU with "reaper" ruid

2018-11-29 Thread George Dunlap
> On Nov 28, 2018, at 6:33 PM, George Dunlap wrote: > >>> -ret = setresuid(dm_uid, dm_uid, 0); >>> +fd = open(lockfile, O_RDWR|O_CREAT, 0666); >>> +if (fd < 0) { >>> +/* All other errno: EBADF, EINVAL, ENOLCK, EWOULDBLOCK */ >>> +LOGED(ERROR,

Re: [Xen-devel] [PATCH 9/9] libxl: Kill QEMU with "reaper" ruid

2018-11-28 Thread George Dunlap
> On Nov 28, 2018, at 5:02 PM, Ian Jackson wrote: > > George Dunlap writes ("[PATCH 9/9] libxl: Kill QEMU with "reaper" ruid"): >> Using kill(-1) to killing an untrusted dm process with the real uid >> equal to the dm_uid isn't guaranteed to succeed: the process in >> question may be able to

Re: [Xen-devel] [PATCH 9/9] libxl: Kill QEMU with "reaper" ruid

2018-11-28 Thread Ian Jackson
George Dunlap writes ("[PATCH 9/9] libxl: Kill QEMU with "reaper" ruid"): > Using kill(-1) to killing an untrusted dm process with the real uid > equal to the dm_uid isn't guaranteed to succeed: the process in > question may be able to kill the reaper process after the setresuid() > and before the

[Xen-devel] [PATCH 9/9] libxl: Kill QEMU with "reaper" ruid

2018-11-23 Thread George Dunlap
Using kill(-1) to killing an untrusted dm process with the real uid equal to the dm_uid isn't guaranteed to succeed: the process in question may be able to kill the reaper process after the setresuid() and before the kill(). Instead, set the real uid to the QEMU user for domain 0