Fedora 22. intel-gpu-tools-2.99.917-22.intel20152.x86_64 Integrated Graphics Chipset: Intel(R) HD Graphics 4400 # dmesg | grep MMIO [ 0.182300] hpet0: at MMIO 0xfed00000, IRQs 2, 8, 0, 0, 0, 0, 0, 0 # intel_gpu_top (intel_gpu_top:5103) intel-mmio-CRITICAL: Test assertion failure function intel_mmio_use_pci_bar, file intel_mmio.c:144: Failed assertion: !(error != 0) Last errno: 1, Operation not permitted Couldn't map MMIO region Stack trace: #0 [_init+0x7eb9] #1 [_init+0x5d96] #2 [_init+0x1291] #3 [__libc_start_main+0xf0] #4 [_init+0x2d11] #5 [<unknown>+0x2d11] Test (null) failed. **** DEBUG **** (intel_gpu_top:5103) intel-mmio-CRITICAL: Test assertion failure function intel_mmio_use_pci_bar, file intel_mmio.c:144: Failed assertion: !(error != 0) Last errno: 1, Operation not permitted Couldn't map MMIO region **** END ****
I presume by the '#' prompt you're running as root, but perhaps it's some selinux policy thing...? Can you try without?
Originally ran it as a user, but tried root because of the permission warning. In any event, same result as user: $ intel_gpu_top (intel_gpu_top:16575) intel-mmio-CRITICAL: Test assertion failure function intel_mmio_use_pci_bar, file intel_mmio.c:144: Failed assertion: !(error != 0) Last errno: 13, Permission denied Couldn't map MMIO region Stack trace: #0 [_init+0x7eb9] #1 [_init+0x5d96] #2 [_init+0x1291] #3 [__libc_start_main+0xf0] #4 [_init+0x2d11] #5 [<unknown>+0x2d11] Test (null) failed. **** DEBUG **** (intel_gpu_top:16575) intel-mmio-CRITICAL: Test assertion failure function intel_mmio_use_pci_bar, file intel_mmio.c:144: Failed assertion: !(error != 0) Last errno: 13, Permission denied Couldn't map MMIO region **** END **** FAIL (84228.913s)
Also, tried setting selinux to permissive. Same result. # setenforce 0 # sestatus SELinux status: enabled SELinuxfs mount: /sys/fs/selinux SELinux root directory: /etc/selinux Loaded policy name: targeted Current mode: permissive Mode from config file: enforcing Policy MLS status: enabled Policy deny_unknown status: allowed Max kernel policy version: 29
(In reply to sean darcy from comment #2) > Originally ran it as a user, but tried root because of the permission > warning. In any event, same result as user: It needs to be run as root. (In reply to sean darcy from comment #3) > Also, tried setting selinux to permissive. Same result. And after that, did you also try as root? I eyeballed libpciaccess sources a bit, but I'm not sure where it could reasonably return EPERM. For the fun of it, what does this say: ls -l /sys/bus/pci/devices/0000:00:02.0/resource*
set selinux to permissive , and ran igt as root: # setenforce 0 # intel_gpu_top (intel_gpu_top:18404) intel-mmio-CRITICAL: Test assertion failure function intel_mmio_use_pci_bar, file intel_mmio.c:144: Failed assertion: !(error != 0) Last errno: 1, Operation not permitted Couldn't map MMIO region Stack trace: #0 [_init+0x7eb9] #1 [_init+0x5d96] #2 [_init+0x1291] #3 [__libc_start_main+0xf0] #4 [_init+0x2d11] #5 [<unknown>+0x2d11] Test (null) failed. **** DEBUG **** (intel_gpu_top:18404) intel-mmio-CRITICAL: Test assertion failure function intel_mmio_use_pci_bar, file intel_mmio.c:144: Failed assertion: !(error != 0) Last errno: 1, Operation not permitted Couldn't map MMIO region **** END **** # ls -l /sys/bus/pci/devices/0000:00:02.0/resource* -r--r--r--. 1 root root 4096 Dec 2 15:09 /sys/bus/pci/devices/0000:00:02.0/resource -rw-------. 1 root root 4194304 Dec 2 15:09 /sys/bus/pci/devices/0000:00:02.0/resource0 -rw-------. 1 root root 268435456 Dec 2 15:09 /sys/bus/pci/devices/0000:00:02.0/resource2 -rw-------. 1 root root 268435456 Dec 2 15:09 /sys/bus/pci/devices/0000:00:02.0/resource2_wc -rw-------. 1 root root 64 Dec 2 15:09 /sys/bus/pci/devices/0000:00:02.0/resource4
I have this distinct feeling someone must have faced this before and knows the answer... anyway, please attach the output of: # strace intel_gpu_top
Created attachment 120302 [details] strace igt
I searched before posting. The only thing I found was a report on fedora: https://bugzilla.redhat.com/show_bug.cgi?id=1281349
messing around, I set vm.mmap_min_addr to zero. #sysctl vm.mmap_min_addr vm.mmap_min_addr = 65536 # sysctl -w vm.mmap_min_addr="0" vm.mmap_min_addr = 0 No joy. Still failed.
I do not know why mmap would return EPERM in this case. open("/sys/bus/pci/devices/0000:00:02.0/resource0", O_RDWR|O_CLOEXEC) = 4 mmap(NULL, 2097152, PROT_READ|PROT_WRITE, MAP_SHARED, 4, 0) = -1 EPERM (Operation not permitted)
Yes, very puzzling. It doesn't seem a corner case. This is a standard vanilla HSW lenovo laptop. Intel(R) Core(TM) i7-4510U CPU @ 2.00GHz kernel-4.2.6-201.fc22.x86_64 Surprised (amazed ?) this hasn't happened to others.
Same issue with 2015Q4 release: intel-gpu-tools-2.99.917-19.20151206.fc23.x86_64 kernel-4.2.8-300.fc23.x86_64 Maybe kernel 4.3 required ?
(In reply to sean darcy from comment #12) > Maybe kernel 4.3 required ? Shouldn't be. My assumption is it's something about your (distro) environment that we haven't thought of.
No reports of anyone else experiencing this, and I think it's safe to say we're not working on this. Sorry, worksforme.
I am having exactly this problem. Fedora 24, 64 bit on a Dell m3800 laptop. I'm wondering if the driver isn't blocking the mmap of PCI resource0 because it is somehow in use. Things that I think could be making a difference: - Optimus setup with Nvidia secondary GPU - DRI 3 is forced on for Intel. - X.org is in use - resource0 is 4 MB in size but the failed mmap is only for 2 MB.
I am having the same problem with intel_panel_fitter on fresh installation of ubuntu mate 16.04 . With Linux Mint 18 it had worked on the same computer well but I had some other problems and had to replace it by ubuntu. $ sudo intel_panel_fitter -p A -x 1200 -y 680 WARNING: This tool is a workaround for people that don't have a Kernel with overscan compensation properties: it is just a temporary solution that may or may not work. Use it at your own risk. (intel_panel_fitter:4194) intel-mmio-CRITICAL: Test assertion failure function intel_mmio_use_pci_bar, file ../../lib/intel_mmio.c:145: (intel_panel_fitter:4194) intel-mmio-CRITICAL: Failed assertion: !(error != 0) (intel_panel_fitter:4194) intel-mmio-CRITICAL: Last errno: 1, Operation not permitted (intel_panel_fitter:4194) intel-mmio-CRITICAL: Couldn't map MMIO region Stack trace: #0 [_init+0x38d9] #1 [_init+0x1536] #2 [_init+0x15e1] #3 [_init+0xa27] #4 [__libc_start_main+0xf0] #5 [_init+0x11a1] #6 [<unknown>+0x11a1] Test (null) failed. **** DEBUG **** (intel_panel_fitter:4194) intel-mmio-CRITICAL: Test assertion failure function intel_mmio_use_pci_bar, file ../../lib/intel_mmio.c:145: (intel_panel_fitter:4194) intel-mmio-CRITICAL: Failed assertion: !(error != 0) (intel_panel_fitter:4194) intel-mmio-CRITICAL: Last errno: 1, Operation not permitted (intel_panel_fitter:4194) intel-mmio-CRITICAL: Couldn't map MMIO region **** END ****
computer description: Desktop Computer product: VM42 (SKU) vendor: ASUSTeK COMPUTER INC. capabilities: smbios-2.7 dmi-2.7 vsyscall32 configuration: administrator_password=disabled boot=normal chassis=desktop family=Vivo PC frontpanel_password=disabled keyboard_password=disabled power-on_password=disabled sku=SKU uuid=[REMOVED] *-core description: Motherboard product: VM42 vendor: ASUSTeK COMPUTER INC. physical id: 0 version: To be filled by O.E.M. serial: [REMOVED] slot: To be filled by O.E.M. *-firmware description: BIOS vendor: ASUSTeK COMPUTER INC. (Licensed from AMI) physical id: 0 version: 1203 date: 08/28/2015 size: 64KiB capacity: 8128KiB capabilities: apm upgrade shadowing cdboot bootselect socketedrom edd int5printscreen int9keyboard acpi usb biosbootspecification uefi *-cpu description: CPU product: Intel(R) Celeron(R) 2957U @ 1.40GHz vendor: Intel Corp. *-memory description: System Memory physical id: 2f slot: System board or motherboard size: 4GiB *-bank:0 description: SODIMM DDR3 Synchronous 1600 MHz (0,6 ns) product: M471B5173EB0-YK0 vendor: Samsung physical id: 0 serial: [REMOVED] slot: ChannelA-DIMM0 size: 4GiB width: 64 bits clock: 1600MHz (0.6ns) *-pci description: Host bridge product: Haswell-ULT DRAM Controller vendor: Intel Corporation physical id: 100 bus info: pci@0000:00:00.0 version: 0b width: 32 bits clock: 33MHz configuration: driver=hsw_uncore resources: irq:0 *-display description: VGA compatible controller product: Haswell-ULT Integrated Graphics Controller vendor: Intel Corporation physical id: 2 bus info: pci@0000:00:02.0 version: 0b width: 64 bits clock: 33MHz capabilities: msi pm vga_controller bus_master cap_list rom configuration: driver=i915 latency=0 resources: irq:43 memory:f7800000-f7bfffff memory:e0000000-efffffff ioport:f000(size=64) *-multimedia:0 description: Audio device product: Haswell-ULT HD Audio Controller vendor: Intel Corporation physical id: 3 bus info: pci@0000:00:03.0 version: 0b width: 64 bits clock: 33MHz capabilities: pm msi pciexpress bus_master cap_list configuration: driver=snd_hda_intel latency=0 resources: irq:46 memory:f7e14000-f7e17fff
$ apt-show-versions intel-gpu-tools intel-gpu-tools:amd64/xenial 1.14-1 uptodate
Does it work from a root shell?
$ sudo -i [sudo] password for dimitry: root@dimitry-VM42:~# intel_panel_fitter -p A -x 1200 -y 680 WARNING: This tool is a workaround for people that don't have a Kernel with overscan compensation properties: it is just a temporary solution that may or may not work. Use it at your own risk. (intel_panel_fitter:4479) intel-mmio-CRITICAL: Test assertion failure function intel_mmio_use_pci_bar, file ../../lib/intel_mmio.c:145: (intel_panel_fitter:4479) intel-mmio-CRITICAL: Failed assertion: !(error != 0) (intel_panel_fitter:4479) intel-mmio-CRITICAL: Last errno: 1, Operation not permitted (intel_panel_fitter:4479) intel-mmio-CRITICAL: Couldn't map MMIO region Stack trace: #0 [_init+0x38d9] #1 [_init+0x1536] #2 [_init+0x15e1] #3 [_init+0xa27] #4 [__libc_start_main+0xf0] #5 [_init+0x11a1] #6 [<unknown>+0x11a1] Test (null) failed. **** DEBUG **** (intel_panel_fitter:4479) intel-mmio-CRITICAL: Test assertion failure function intel_mmio_use_pci_bar, file ../../lib/intel_mmio.c:145: (intel_panel_fitter:4479) intel-mmio-CRITICAL: Failed assertion: !(error != 0) (intel_panel_fitter:4479) intel-mmio-CRITICAL: Last errno: 1, Operation not permitted (intel_panel_fitter:4479) intel-mmio-CRITICAL: Couldn't map MMIO region **** END **** FAIL (-1.000s) root@dimitry-VM42:~#
As you can see using a root shell does not help either. If you need any other information please let me know. Regards, Dimitry
user@Lenovo-G510:~$ apt-show-versions intel-gpu-tools intel-gpu-tools:amd64/xenial 1.14-1 uptodate user@Lenovo-G510:~$ sudo intel_gpu_top [sudo] password for olokki: (intel_gpu_top:24757) intel-mmio-CRITICAL: Test assertion failure function intel_mmio_use_pci_bar, file ../../lib/intel_mmio.c:145: (intel_gpu_top:24757) intel-mmio-CRITICAL: Failed assertion: !(error != 0) (intel_gpu_top:24757) intel-mmio-CRITICAL: Last errno: 1, Operation not permitted (intel_gpu_top:24757) intel-mmio-CRITICAL: Couldn't map MMIO region Stack trace: #0 [_init+0x83d9] #1 [_init+0x6036] #2 [_init+0x1301] #3 [__libc_start_main+0xf0] #4 [_init+0x2f61] #5 [<unknown>+0x2f61] Test (null) failed. **** DEBUG **** (intel_gpu_top:24757) intel-mmio-CRITICAL: Test assertion failure function intel_mmio_use_pci_bar, file ../../lib/intel_mmio.c:145: (intel_gpu_top:24757) intel-mmio-CRITICAL: Failed assertion: !(error != 0) (intel_gpu_top:24757) intel-mmio-CRITICAL: Last errno: 1, Operation not permitted (intel_gpu_top:24757) intel-mmio-CRITICAL: Couldn't map MMIO region **** END **** FAIL (-1.000s)
Moving this from DRM/Intel to IGT, since it seems to be problem with that. Sean, Jonathan, Dimitry and Rami - please check if this problem is still valid with the latest igt.
(In reply to Jari Tahvanainen from comment #23) > Moving this from DRM/Intel to IGT, since it seems to be problem with that. > Sean, Jonathan, Dimitry and Rami - please check if this problem is still > valid with the latest igt. I've found i have the same issue, running as root: root@oskar-ThinkPad-X270:/sys/bus/pci/devices/0000:00:02.0# intel_gpu_top (intel_gpu_top:6971) intel-mmio-CRITICAL: Test assertion failure function intel_mmio_use_pci_bar, file ../../lib/intel_mmio.c:145: (intel_gpu_top:6971) intel-mmio-CRITICAL: Failed assertion: !(error != 0) (intel_gpu_top:6971) intel-mmio-CRITICAL: Last errno: 1, Operation not permitted (intel_gpu_top:6971) intel-mmio-CRITICAL: Couldn't map MMIO region Stack trace: #0 [_init+0x7369] #1 [_init+0x4fd0] #2 [_init+0x98e] #3 [__libc_start_main+0xf1] #4 [_init+0x1752] #5 [<unknown>+0x1752] Test (null) failed. **** DEBUG **** (intel_gpu_top:6971) intel-chipset-DEBUG: Test requirement passed: pci_dev (intel_gpu_top:6971) intel-mmio-CRITICAL: Test assertion failure function intel_mmio_use_pci_bar, file ../../lib/intel_mmio.c:145: (intel_gpu_top:6971) intel-mmio-CRITICAL: Failed assertion: !(error != 0) (intel_gpu_top:6971) intel-mmio-CRITICAL: Last errno: 1, Operation not permitted (intel_gpu_top:6971) intel-mmio-CRITICAL: Couldn't map MMIO region **** END **** FAIL (-1.000s) strace shows failure on opening resource0: open("/sys/bus/pci/devices/0000:00:02.0/resource", O_RDONLY|O_CLOEXEC) = 4 read(4, "0x00000000e0000000 0x00000000e0f"..., 512) = 512 close(4) = 0 open("/sys/bus/pci/devices/0000:00:02.0/resource0", O_RDWR|O_CLOEXEC) = 4 mmap(NULL, 2097152, PROT_READ|PROT_WRITE, MAP_SHARED, 4, 0) = -1 EPERM (Operation not permitted) close(4) = 0 write(2, "(intel_gpu_top:7094) intel-mmio-"..., 131(intel_gpu_top:7094) intel-mmio-CRITICAL: Test assertion failure function intel_mmio_use_pci_bar, file ../../lib/intel_mmio.c:145: ) = 131 write(2, "(intel_gpu_top:7094) intel-mmio-"..., 74(intel_gpu_top:7094) intel-mmio-CRITICAL: Failed assertion: !(error != 0) ) = 74 write(2, "(intel_gpu_top:7094) intel-mmio-"..., 81(intel_gpu_top:7094) intel-mmio-CRITICAL: Last errno: 1, Operation not permitted ) = 81 write(2, "(intel_gpu_top:7094) intel-mmio-"..., 67(intel_gpu_top:7094) intel-mmio-CRITICAL: Couldn't map MMIO region ) = 67 checking if readable with cat: root@oskar-ThinkPad-X270:/sys/bus/pci/devices/0000:00:02.0# cat resource 0x00000000e0000000 0x00000000e0ffffff 0x0000000000140204 0x0000000000000000 0x0000000000000000 0x0000000000000000 0x00000000c0000000 0x00000000dfffffff 0x000000000014220c 0x0000000000000000 0x0000000000000000 0x0000000000000000 0x000000000000e000 0x000000000000e03f 0x0000000000040101 0x0000000000000000 0x0000000000000000 0x0000000000000000 0x00000000000c0000 0x00000000000dffff 0x0000000000000212 0x0000000000000000 0x0000000000000000 0x0000000000000000 0x0000000000000000 0x0000000000000000 0x0000000000000000 0x0000000000000000 0x0000000000000000 0x0000000000000000 0x0000000000000000 0x0000000000000000 0x0000000000000000 0x0000000000000000 0x0000000000000000 0x0000000000000000 0x0000000000000000 0x0000000000000000 0x0000000000000000 root@oskar-ThinkPad-X270:/sys/bus/pci/devices/0000:00:02.0# cat resource0 cat: resource0: Input/output error root@oskar-ThinkPad-X270:/sys/bus/pci/devices/0000:00:02.0# cat resource2 cat: resource2: Input/output error root@oskar-ThinkPad-X270:/sys/bus/pci/devices/0000:00:02.0# cat resource2_wc cat: resource2_wc: Input/output error root@oskar-ThinkPad-X270:/sys/bus/pci/devices/0000:00:02.0# cat resource4 cat: resource4: Invalid argument root@oskar-ThinkPad-X270:/sys/bus/pci/devices/0000:00:02.0# ls -la resource* -r--r--r-- 1 root root 4096 okt 2 22:38 resource -rw------- 1 root root 16777216 okt 2 22:30 resource0 -rw------- 1 root root 536870912 okt 2 22:38 resource2 -rw------- 1 root root 536870912 okt 2 22:38 resource2_wc -rw------- 1 root root 64 okt 2 22:38 resource4 root@oskar-ThinkPad-X270:/sys/bus/pci/devices/0000:00:02.0# uname -a Linux oskar-ThinkPad-X270 4.10.0-35-generic #39-Ubuntu SMP Wed Sep 13 07:46:59 UTC 2017 x86_64 x86_64 x86_64 GNU/Linux I wonder if it could be related to secure_boot?
Forgot to mention this is ubuntu 17.04 and i intel-gpu-tools are version 1.17-1
Just stumbled into the same issue, disabled secure boot, and the issue disappeared. Now I am having a different issue, but I think it is just my version being too old.
Based on an IRC chat with Joonas and Dave, looks like this is indeed kernel lockdown on UEFI secure boot doing its magic. The kernel lockdown patches haven't been merged upstream, so you should not be able to reproduce this on kernel.org upstream kernels. Please file your bugs against your distro if this remains an issue for you. See e.g. https://mjg59.dreamwidth.org/50577.html for background. Sorry it took us so long to get this; we're obviously not running non-upstream kernels with lockdown enabled.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.