<html> <head> <meta http-equiv="Content-Type" content="text/html; charset=utf-8"> <meta name="Generator" content="Microsoft Exchange Server"> <style></style> </head> <body> <div> <div dir="auto" style="direction:ltr; margin:0; padding:0; font-family:sans-serif; font-size:11pt; color:black"> Already fixed on upstream.<br> </div> <div dir="auto" style="direction:ltr; margin:0; padding:0; font-family:sans-serif; font-size:11pt; color:black"> <a href="https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/kernel/locking?h=v4.18-rc4&id=03eeafdd9ab06a770d42c2b264d50dff7e2f4eee">https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/kernel/locking?h=v4.18-rc4&id=03eeafdd9ab06a770d42c2b264d50dff7e2f4eee</a><br> </div> <hr tabindex="-1" style="display:inline-block; width:98%"> <div id="x_divRplyFwdMsg" dir="ltr"><font face="Calibri, sans-serif" color="#000000" style="font-size:11pt"><b>From:</b> Pan, Xinhui<br> <b>Sent:</b> Friday, July 13, 2018 6:17:06 PM<br> <b>To:</b> Kuehling, Felix; Michel Dänzer<br> <b>Cc:</b> amd-gfx@lists.freedesktop.org<br> <b>Subject:</b> RE: DEBUG_LOCKS_WARN_ON(sem->owner != ((struct task_struct *)(1UL << 0)))</font> <div> </div> </div> </div> <font size="2"><span style="font-size:11pt;"> <div class="PlainText">Reproduced on my side. This backtrace only shows one time after boot up unless reboot again.<br> <br> This backtrace is introduced by KFDIPCTest.BasicTest(other tests also can introduce it).<br> This test defines some buffers which are mapped to gpu, and the fork later causes this warning. <br> <br> The code comments says libhfskmt should do cleanup work after fork as these buffers are invalid in child process.<br> But the warning shows during fork(), how libhsakmt do such cleanup work?<br> <br> -----Original Message-----<br> From: Kuehling, Felix <br> Sent: 2018年7月13日 0:32<br> To: Michel Dänzer <michel@daenzer.net>; Pan, Xinhui <Xinhui.Pan@amd.com><br> Cc: amd-gfx@lists.freedesktop.org<br> Subject: Re: DEBUG_LOCKS_WARN_ON(sem->owner != ((struct task_struct *)(1UL << 0)))<br> <br> [+Pan Xinhui]<br> <br> On 2018-07-12 03:16 AM, Michel Dänzer wrote:<br> > On 2018-07-12 02:43 AM, Felix Kuehling wrote:<br> >> Kent just caught a similar backtrace in one of our KFD pre-submission <br> >> tests (see below)<br> > Yeah, looks the same.<br> ><br> ><br> >> Neither KFD nor AMDGPU are implied in the backtrace. Is this a <br> >> regression in the kernel itself? amd-kfd-staging is currently based <br> >> on 4.18-rc1.<br> > FWIW, I saw this with 4.17 based kernels already, and I didn't have <br> > CONFIG_DEBUG_RWSEMS enabled with older kernels, so I'm not sure it's a<br> > (recent) regression.<br> <br> I've now also seen it on Oded's branch (4.17-rc5). It is reproduced reliably by a new test that Pan Xinhui just added to our kfdtest (KFDMemoryTest.MMapLarge).<br> <br> Regards,<br> Felix<br> <br> <br> </div> </span></font> </body> </html>