<html> <head> <meta http-equiv="Content-Type" content="text/html; charset=us-ascii"> <style type="text/css" style="display:none;"> P {margin-top:0;margin-bottom:0;} </style> </head> <body dir="ltr"> <p style="font-family:Arial;font-size:10pt;color:#317100;margin:15pt;" align="Left"> [AMD Public Use]<br> </p> <br> <div> <div style="font-family: Calibri, Arial, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);"> Thanks Lyude for testing the patch.</div> <div style="font-family: Calibri, Arial, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);"> <br> </div> <div style="font-family: Calibri, Arial, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);"> Are you referring to this issue [1] ?</div> <div style="font-family: Calibri, Arial, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);"> <br> </div> <div style="font-family: Calibri, Arial, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);"> Is it reproducible after applying this patch as well ?</div> <div style="font-family: Calibri, Arial, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);"> <br> </div> <div style="font-family: Calibri, Arial, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);"> [1] <a href="https://gitlab.freedesktop.org/drm/amd/-/issues/1473" id="LPlnk">https://gitlab.freedesktop.org/drm/amd/-/issues/1473</a></div> <div> <div style="font-family: Calibri, Arial, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);"> <br> </div> <div id="Signature"> <div> <div style="font-family:Calibri,Arial,Helvetica,sans-serif; font-size:12pt; color:rgb(0,0,0)"> -- Anson</div> </div> </div> </div> <div id="appendonsend"></div> <hr style="display:inline-block;width:98%" tabindex="-1"> <div id="divRplyFwdMsg" dir="ltr"><font face="Calibri, sans-serif" style="font-size:11pt" color="#000000"><b>From:</b> Lyude Paul <lyude@redhat.com><br> <b>Sent:</b> Friday, March 5, 2021 6:08 PM<br> <b>To:</b> Jacob, Anson <Anson.Jacob@amd.com>; amd-gfx@lists.freedesktop.org <amd-gfx@lists.freedesktop.org><br> <b>Cc:</b> Deucher, Alexander <Alexander.Deucher@amd.com>; Kuehling, Felix <Felix.Kuehling@amd.com><br> <b>Subject:</b> Re: [PATCH] drm/amdkfd: Fix UBSAN shift-out-of-bounds warning</font> <div> </div> </div> <div class="BodyFragment"><font size="2"><span style="font-size:11pt;"> <div class="PlainText">Tested-by: Lyude Paul <lyude@redhat.com><br> <br> That just leaves the KASAN error from read_indirect_azalia_reg, thanks for the<br> fix!<br> <br> On Thu, 2021-03-04 at 15:08 -0500, Anson Jacob wrote:<br> > If get_num_sdma_queues or get_num_xgmi_sdma_queues is 0, we end up<br> > doing a shift operation where the number of bits shifted equals<br> > number of bits in the operand. This behaviour is undefined.<br> > <br> > Set num_sdma_queues or num_xgmi_sdma_queues to ULLONG_MAX, if the<br> > count is >= number of bits in the operand.<br> > <br> > Bug: <a href="https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgitlab.freedesktop.org%2Fdrm%2Famd%2F-%2Fissues%2F1472&data=04%7C01%7Canson.jacob%40amd.com%7Ce9ea7130bcd945a194aa08d8e02ba9f8%7C3dd8961fe4884e608e11a82d994e183d%7C0%7C0%7C637505826204096667%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&sdata=GEyObPt%2FnEeU9Y7K2I30RS1TBMGbbUneQ6hWkR7pJLM%3D&reserved=0"> https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgitlab.freedesktop.org%2Fdrm%2Famd%2F-%2Fissues%2F1472&data=04%7C01%7Canson.jacob%40amd.com%7Ce9ea7130bcd945a194aa08d8e02ba9f8%7C3dd8961fe4884e608e11a82d994e183d%7C0%7C0%7C637505826204096667%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&sdata=GEyObPt%2FnEeU9Y7K2I30RS1TBMGbbUneQ6hWkR7pJLM%3D&reserved=0</a><br> > Reported-by: Lyude Paul <lyude@redhat.com><br> > Signed-off-by: Anson Jacob <Anson.Jacob@amd.com><br> > Reviewed-by: Alex Deucher <alexander.deucher@amd.com><br> > Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com><br> > ---<br> > .../drm/amd/amdkfd/kfd_device_queue_manager.c | 17 +++++++++++++++--<br> > 1 file changed, 15 insertions(+), 2 deletions(-)<br> > <br> > diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c<br> > b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c<br> > index c37e9c4b1fb4..e7a3c496237f 100644<br> > --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c<br> > +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c<br> > @@ -1128,6 +1128,9 @@ static int set_sched_resources(struct<br> > device_queue_manager *dqm)<br> > <br> > static int initialize_cpsch(struct device_queue_manager *dqm)<br> > {<br> > + uint64_t num_sdma_queues;<br> > + uint64_t num_xgmi_sdma_queues;<br> > +<br> > pr_debug("num of pipes: %d\n", get_pipes_per_mec(dqm));<br> > <br> > mutex_init(&dqm->lock_hidden);<br> > @@ -1136,8 +1139,18 @@ static int initialize_cpsch(struct device_queue_manager<br> > *dqm)<br> > dqm->active_cp_queue_count = 0;<br> > dqm->gws_queue_count = 0;<br> > dqm->active_runlist = false;<br> > - dqm->sdma_bitmap = ~0ULL >> (64 - get_num_sdma_queues(dqm));<br> > - dqm->xgmi_sdma_bitmap = ~0ULL >> (64 - get_num_xgmi_sdma_queues(dqm));<br> > +<br> > + num_sdma_queues = get_num_sdma_queues(dqm);<br> > + if (num_sdma_queues >= BITS_PER_TYPE(dqm->sdma_bitmap))<br> > + dqm->sdma_bitmap = ULLONG_MAX;<br> > + else<br> > + dqm->sdma_bitmap = (BIT_ULL(num_sdma_queues) - 1);<br> > +<br> > + num_xgmi_sdma_queues = get_num_xgmi_sdma_queues(dqm);<br> > + if (num_xgmi_sdma_queues >= BITS_PER_TYPE(dqm->xgmi_sdma_bitmap))<br> > + dqm->xgmi_sdma_bitmap = ULLONG_MAX;<br> > + else<br> > + dqm->xgmi_sdma_bitmap = (BIT_ULL(num_xgmi_sdma_queues) - 1);<br> > <br> > INIT_WORK(&dqm->hw_exception_work, kfd_process_hw_exception);<br> > <br> <br> -- <br> Sincerely,<br> Lyude Paul (she/her)<br> Software Engineer at Red Hat<br> <br> Note: I deal with a lot of emails and have a lot of bugs on my plate. If you've<br> asked me a question, are waiting for a review/merge on a patch, etc. and I<br> haven't responded in a while, please feel free to send me another email to check<br> on my status. I don't bite!<br> <br> </div> </span></font></div> </div> </body> </html>