Hi, recently we’ve seen that machines that have Nvidia GPUs running the latest driver will crash upon running Simics (4.6) (otherwise X starts up and runs fine). Other machines without Nvidia cards but running the same OS/patch level run fine. This appears to have only begun a few weeks ago, so I’m guessing it started with the 410.7x driver that would have been released around then. Looking at the X.org log appears to confirm the driver’s at fault:
[ 20778.682] (–) NVIDIA(GPU-0):
[ 21833.428] (EE)
[ 21833.428] (EE) Backtrace:
[ 21833.428] (EE) 0: /usr/bin/X (xorg_backtrace+0x55) [0x55bc45235185]
[ 21833.428] (EE) 1: /usr/bin/X (0x55bc45084000+0x1b4e09) [0x55bc45238e09]
[ 21833.428] (EE) 2: /lib64/libpthread.so.0 (0x7efcbc93e000+0xf5d0) [0x7efcbc94d5d0]
[ 21833.428] (EE) 3: /usr/lib64/xorg/modules/drivers/nvidia_drv.so (0x7efcb82b4000+0xbab90) [0x7efcb836eb90]
[ 21833.428] (EE) 4: /usr/lib64/xorg/modules/libwfb.so (wfbSolid+0x331) [0x7efcb7e93a51]
[ 21833.428] (EE) 5: /usr/lib64/xorg/modules/libwfb.so (wfbFill+0x50f) [0x7efcb7e8afcf]
[ 21833.428] (EE) 6: /usr/lib64/xorg/modules/libwfb.so (wfbPolyFillRect+0x1a0) [0x7efcb7e8bb20]
[ 21833.428] (EE) 7: /usr/lib64/xorg/modules/drivers/nvidia_drv.so (0x7efcb82b4000+0x645002) [0x7efcb88f9002]
[ 21833.428] (EE)
[ 21833.428] (EE) Segmentation fault at address 0xffff98fa400fa00c
[ 21833.428] (EE)
Fatal server error:
[ 21833.428] (EE) Caught signal 11 (Segmentation fault). Server aborting
[ 21833.428] (EE)
[ 21833.428] (EE)
These are running RHEL 7.6 with the 3.10.0-957 kernel. The upgrade to 7.6 was recent though, the nvidia problem showed up when we were still on 7.5.
dmesg has a lot of this (not sure it’s relevant)
[20719.347191] nvidia 0000:01:00.0: irq 136 for MSI/MSI-X
nvidia-bug-report.log.gz (174 KB)