GPU hang on Cherry Trail

[johalun]

[nomadlogic]

hey there - were you able to capture a core from this panic? if so it may be helpful to post the full backtrace in this issue.

[johalun]

Sorry no core. A second after this output the system automatically reboots and no core or anything remains..

[johalun]

I could get a core. It seems you need a swap partition for that.. Since i run on USB memory I deactivated swap...

(kgdb) bt
#0  doadump (textdump=1) at pcpu.h:221
#1  0xffffffff80a409e5 in kern_reboot (howto=<value optimized out>) at /home/mirama/dev/freebsd-base-graphics/sys/kern/kern_shutdown.c:366
#2  0xffffffff80a40fbb in vpanic (fmt=<value optimized out>, ap=<value optimized out>) at /home/mirama/dev/freebsd-base-graphics/sys/kern/kern_shutdown.c:767
#3  0xffffffff80a41003 in panic (fmt=0x0) at /home/mirama/dev/freebsd-base-graphics/sys/kern/kern_shutdown.c:690
#4  0xffffffff80eaf461 in trap_fatal (frame=0xfffffe0048f5d0d0, eva=16) at /home/mirama/dev/freebsd-base-graphics/sys/amd64/amd64/trap.c:841
#5  0xffffffff80eaf66d in trap_pfault (frame=0xfffffe0048f5d0d0, usermode=0) at /home/mirama/dev/freebsd-base-graphics/sys/amd64/amd64/trap.c:691
#6  0xffffffff80eaeb54 in trap (frame=0xfffffe0048f5d0d0) at /home/mirama/dev/freebsd-base-graphics/sys/amd64/amd64/trap.c:442
#7  0xffffffff80e8ef31 in calltrap () at /home/mirama/dev/freebsd-base-graphics/sys/amd64/amd64/exception.S:236
#8  0xffffffff82ba1b19 in pci_dev_put (pdev=0x0) at /home/mirama/dev/freebsd-base-graphics/sys/modules/linuxkpi/../../compat/linuxkpi/common/src/linux_pci.c:386
#9  0xffffffff82a0ab37 in intel_detect_pch (dev=<value optimized out>)
    at /home/mirama/dev/freebsd-base-graphics/sys/modules/drm2/i915/i915kms/../../../../dev/drm2/i915/i915_drv.c:522
#10 0xffffffff82a0913c in i915_driver_load (dev=0xfffff80006903000, flags=<value optimized out>)
    at /home/mirama/dev/freebsd-base-graphics/sys/modules/drm2/i915/i915kms/../../../../dev/drm2/i915/i915_dma.c:1048
#11 0xffffffff82b38555 in drm_dev_register (dev=0xfffff80006903000, flags=18446744071606949548)
    at /home/mirama/dev/freebsd-base-graphics/sys/modules/drm2/drm2/../../../dev/drm2/drm_drv.c:785
#12 0xffffffff82b518f9 in drm_get_pci_dev (pdev=0xfffff80003a1f000, ent=0xffffffff82acc5d0, driver=<value optimized out>)
    at /home/mirama/dev/freebsd-base-graphics/sys/modules/drm2/drm2/../../../dev/drm2/drm_pci.c:323
#13 0xffffffff82ba1f83 in linux_pci_attach (dev=<value optimized out>)
    at /home/mirama/dev/freebsd-base-graphics/sys/modules/linuxkpi/../../compat/linuxkpi/common/src/linux_pci.c:193
#14 0xffffffff80a750f0 in device_attach (dev=0xfffff8000395d000) at device_if.h:180
#15 0xffffffff80a767d6 in bus_generic_driver_added (dev=<value optimized out>, driver=<value optimized out>)
    at /home/mirama/dev/freebsd-base-graphics/sys/kern/subr_bus.c:2858
#16 0xffffffff80a72abd in devclass_driver_added (dc=<value optimized out>, driver=<value optimized out>) at bus_if.h:204
#17 0xffffffff80a729e1 in devclass_add_driver (dc=<value optimized out>, driver=<value optimized out>, pass=<value optimized out>, dcp=<value optimized out>)
    at /home/mirama/dev/freebsd-base-graphics/sys/kern/subr_bus.c:1172
#18 0xffffffff82ba1816 in pci_register_driver (pdrv=<value optimized out>)
    at /home/mirama/dev/freebsd-base-graphics/sys/modules/linuxkpi/../../compat/linuxkpi/common/src/linux_pci.c:297
#19 0xffffffff82a0cd7c in _module_run (arg=<value optimized out>) at module.h:80
#20 0xffffffff80a14478 in linker_load_module (kldname=<value optimized out>, modname=0xfffff80003ed7800 "i915kms", parent=<value optimized out>, 
    verinfo=<value optimized out>, lfpp=<value optimized out>) at /home/mirama/dev/freebsd-base-graphics/sys/kern/kern_linker.c:230
#21 0xffffffff80a15ad7 in kern_kldload (td=<value optimized out>, file=<value optimized out>, fileid=0xfffffe0048f5dac4)
    at /home/mirama/dev/freebsd-base-graphics/sys/kern/kern_linker.c:1037
#22 0xffffffff80a15b9b in sys_kldload (td=0xfffff8000651e000, uap=<value optimized out>) at /home/mirama/dev/freebsd-base-graphics/sys/kern/kern_linker.c:1063
#23 0xffffffff80eafc1b in amd64_syscall (td=0xfffff8000651e000, traced=0) at subr_syscall.c:135
#24 0xffffffff80e8f21b in Xfast_syscall () at /home/mirama/dev/freebsd-base-graphics/sys/amd64/amd64/exception.S:396
#25 0x000000080086d12a in ?? ()

[mattmacy]

Please try the latest.

[johalun]

Got a bit further this time.

(kgdb) bt
#0  doadump (textdump=1) at pcpu.h:221
#1  0xffffffff80a40b85 in kern_reboot (howto=<value optimized out>) at /usr/home/johannes/dev/freebsd/freebsd-base-graphics/sys/kern/kern_shutdown.c:366
#2  0xffffffff80a4115b in vpanic (fmt=<value optimized out>, ap=<value optimized out>)
    at /usr/home/johannes/dev/freebsd/freebsd-base-graphics/sys/kern/kern_shutdown.c:767
#3  0xffffffff80a411a3 in panic (fmt=0x0) at /usr/home/johannes/dev/freebsd/freebsd-base-graphics/sys/kern/kern_shutdown.c:690
#4  0xffffffff80eaf401 in trap_fatal (frame=0xfffffe0048ef0fb0, eva=0) at /usr/home/johannes/dev/freebsd/freebsd-base-graphics/sys/amd64/amd64/trap.c:841
#5  0xffffffff80eaf090 in trap (frame=0xfffffe0048ef0fb0) at /usr/home/johannes/dev/freebsd/freebsd-base-graphics/sys/amd64/amd64/trap.c:203
#6  0xffffffff80e8f881 in calltrap () at /usr/home/johannes/dev/freebsd/freebsd-base-graphics/sys/amd64/amd64/exception.S:236
#7  0xffffffff82b24ef0 in drm_clflush_virt_range (addr=0xfffff8000a70b000, length=<value optimized out>)
    at /usr/home/johannes/dev/freebsd/freebsd-base-graphics/sys/modules/drm2/drm2/../../../dev/drm2/drm_cache.c:139
#8  0xffffffff82a21271 in __hw_ppgtt_init (dev=<value optimized out>, ppgtt=<value optimized out>)
    at /usr/home/johannes/dev/freebsd/freebsd-base-graphics/sys/modules/drm2/i915/i915kms/../../../../dev/drm2/i915/i915_gem_gtt.c:362
#9  0xffffffff82a21cac in i915_ppgtt_create (dev=0xfffff80003f9c000, fpriv=0x0)
    at /usr/home/johannes/dev/freebsd/freebsd-base-graphics/sys/modules/drm2/i915/i915kms/../../../../dev/drm2/i915/i915_gem_gtt.c:2159
#10 0xffffffff82a19978 in i915_gem_create_context (dev=<value optimized out>, file_priv=<value optimized out>)
    at /usr/home/johannes/dev/freebsd/freebsd-base-graphics/sys/modules/drm2/i915/i915kms/../../../../dev/drm2/i915/i915_gem_context.c:299
#11 0xffffffff82a19696 in i915_gem_context_init (dev=0xfffff80003f9c000)
    at /usr/home/johannes/dev/freebsd/freebsd-base-graphics/sys/modules/drm2/i915/i915kms/../../../../dev/drm2/i915/i915_gem_context.c:391
#12 0xffffffff82a16cc0 in i915_gem_init (dev=0xfffff80003f9c000)
    at /usr/home/johannes/dev/freebsd/freebsd-base-graphics/sys/modules/drm2/i915/i915kms/../../../../dev/drm2/i915/i915_gem.c:5039
#13 0xffffffff82a09b25 in i915_driver_load (dev=<value optimized out>, flags=<value optimized out>)
    at /usr/home/johannes/dev/freebsd/freebsd-base-graphics/sys/modules/drm2/i915/i915kms/../../../../dev/drm2/i915/i915_dma.c:414
#14 0xffffffff82b3b4d5 in drm_dev_register (dev=0xfffff80003f9c000, flags=18446744071606956220)
    at /usr/home/johannes/dev/freebsd/freebsd-base-graphics/sys/modules/drm2/drm2/../../../dev/drm2/drm_drv.c:785
#15 0xffffffff82b55de9 in drm_prime_pages_to_sg (pages=<value optimized out>, nr_pages=<value optimized out>) at scatterlist.h:110
#16 0xffffffff82ba7839 in linux_pci_attach (dev=<value optimized out>)
    at /usr/home/johannes/dev/freebsd/freebsd-base-graphics/sys/modules/linuxkpi/../../compat/linuxkpi/common/src/linux_pci.c:210
#17 0xffffffff80a75320 in device_attach (dev=0xfffff8000393a600) at device_if.h:180
#18 0xffffffff80a76a06 in bus_generic_driver_added (dev=<value optimized out>, driver=<value optimized out>)
    at /usr/home/johannes/dev/freebsd/freebsd-base-graphics/sys/kern/subr_bus.c:2858
#19 0xffffffff80a72ced in devclass_driver_added (dc=<value optimized out>, driver=<value optimized out>) at bus_if.h:204
#20 0xffffffff80a72c11 in devclass_add_driver (dc=<value optimized out>, driver=<value optimized out>, pass=<value optimized out>, dcp=<value optimized out>)
    at /usr/home/johannes/dev/freebsd/freebsd-base-graphics/sys/kern/subr_bus.c:1172
#21 0xffffffff82ba704f in pci_register_driver (pdrv=<value optimized out>)
    at /usr/home/johannes/dev/freebsd/freebsd-base-graphics/sys/modules/linuxkpi/../../compat/linuxkpi/common/src/linux_pci.c:327
#22 0xffffffff82a0cd2c in _module_run (arg=<value optimized out>) at module.h:80
#23 0xffffffff80a14618 in linker_load_module (kldname=<value optimized out>, modname=0xfffff800039f3000 "i915kms", parent=<value optimized out>, 
    verinfo=<value optimized out>, lfpp=<value optimized out>) at /usr/home/johannes/dev/freebsd/freebsd-base-graphics/sys/kern/kern_linker.c:230
---Type <return> to continue, or q <return> to quit---
#24 0xffffffff80a15c77 in kern_kldload (td=<value optimized out>, file=<value optimized out>, fileid=0xfffffe0048ef1ac4)
    at /usr/home/johannes/dev/freebsd/freebsd-base-graphics/sys/kern/kern_linker.c:1037
#25 0xffffffff80a15d3b in sys_kldload (td=0xfffff80003ef0000, uap=<value optimized out>)
    at /usr/home/johannes/dev/freebsd/freebsd-base-graphics/sys/kern/kern_linker.c:1063
#26 0xffffffff80eafbbb in amd64_syscall (td=0xfffff80003ef0000, traced=0) at subr_syscall.c:135
#27 0xffffffff80e8fb6b in Xfast_syscall () at /usr/home/johannes/dev/freebsd/freebsd-base-graphics/sys/amd64/amd64/exception.S:396
#28 0x000000080086d12a in ?? ()
Previous frame inner to this frame (corrupt stack?)
Current language:  auto; currently minimal

[mattmacy]

It looks like it may be trying to use clflushopt on an unsupported processor. Can you try the latest and see if it works? If not I'll just use clflush.

Thanks.

[johalun]

Getting closer :)

[mattmacy]

Hardly shippable, but this is definitely progress.

[johalun]

It actually renders something when I start X but the screen content is all messed up.. Like tiles repeating in X and Y.

[mattmacy]

What does the log show if you set dev.drm.drm_debug=-1 (after loading i915kms) before starting X?

[johalun]

Here's the log. From boot to kldload i915kms and start/stop X a couple of times.

messages.txt.zip

[mattmacy]

This is returning EIO. I'll have to dig in to which path is doing that.

/* Throttle our rendering by waiting until the ring has completed our requests
 * emitted over 20 msec ago.
 *
 * Note that if we were to use the current jiffies each time around the loop,
 * we wouldn't escape the function with any frames outstanding if the time to
 * render a frame was over 20ms.
 *
 * This should get us reasonable parallelism between CPU and GPU but also
 * relatively low latency when blocking on a particular request to finish.
 */
static int
i915_gem_ring_throttle(struct drm_device *dev, struct drm_file *file)
{
    struct drm_i915_private *dev_priv = dev->dev_private;
    struct drm_i915_file_private *file_priv = file->driver_priv;
    unsigned long recent_enough = jiffies - DRM_I915_THROTTLE_JIFFIES;
    struct drm_i915_gem_request *request, *target = NULL;
    unsigned reset_counter;
    int ret;

    ret = i915_gem_wait_for_error(&dev_priv->gpu_error);
    if (ret)
        return ret;

    ret = i915_gem_check_wedge(&dev_priv->gpu_error, false);
    if (ret)
        return ret;

    spin_lock(&file_priv->mm.lock);
    list_for_each_entry(request, &file_priv->mm.request_list, client_list) {
        if (time_after_eq(request->emitted_jiffies, recent_enough))
            break;

        /*
         * Note that the request might not have been submitted yet.
         * In which case emitted_jiffies will be zero.
         */
        if (!request->emitted_jiffies)
            continue;

        target = request;
    }
    reset_counter = atomic_read(&dev_priv->gpu_error.reset_counter);
    if (target)
        i915_gem_request_reference(target);
    spin_unlock(&file_priv->mm.lock);

    if (target == NULL)
        return 0;

    ret = __i915_wait_request(target, reset_counter, true, NULL, NULL);
    if (ret == 0)
        queue_delayed_work(dev_priv->wq, &dev_priv->mm.retire_work, 0);

    i915_gem_request_unreference__unlocked(target);

    return ret;
}

[johalun]

I will add some printf's and test.

[johalun]

Ok. I get the many returns at
ret = i915_gem_check_wedge(&dev_priv->gpu_error, false);

[mattmacy]

Which one though?

int
i915_gem_check_wedge(struct i915_gpu_error *error,
             bool interruptible)
{
    if (i915_reset_in_progress(error)) {
        /* Non-interruptible callers can't handle -EAGAIN, hence return
         * -EIO unconditionally for these. */
        if (!interruptible)
            return -EIO;

        /* Recovery complete, but the reset failed ... */
        if (i915_terminally_wedged(error))
            return -EIO;

        /*
         * Check if GPU Reset is in progress - we need intel_ring_begin
         * to work properly to reinit the hw state while the gpu is
         * still marked as reset-in-progress. Handle this with a flag.
         */
        if (!error->reload_in_reset)
            return -EAGAIN;
    }

    return 0;
}

[mattmacy]

/**
 * i915_reset_and_wakeup - do process context error handling work
 * @dev: drm device
 *
 * Fire an error uevent so userspace can see that a hang or error
 * was detected.
 */
static void i915_reset_and_wakeup(struct drm_device *dev)
{
    struct drm_i915_private *dev_priv = to_i915(dev);
    struct i915_gpu_error *error = &dev_priv->gpu_error;
    char *error_event[] = { I915_ERROR_UEVENT "=1", NULL };
    char *reset_event[] = { I915_RESET_UEVENT "=1", NULL };
    char *reset_done_event[] = { I915_ERROR_UEVENT "=0", NULL };
    int ret;

    kobject_uevent_env(&dev->primary->kdev->kobj, KOBJ_CHANGE, error_event);

    /*
     * Note that there's only one work item which does gpu resets, so we
     * need not worry about concurrent gpu resets potentially incrementing
     * error->reset_counter twice. We only need to take care of another
     * racing irq/hangcheck declaring the gpu dead for a second time. A
     * quick check for that is good enough: schedule_work ensures the
     * correct ordering between hang detection and this work item, and since
     * the reset in-progress bit is only ever set by code outside of this
     * work we don't need to worry about any other races.
     */
    if (i915_reset_in_progress(error) && !i915_terminally_wedged(error)) {
        DRM_DEBUG_DRIVER("resetting chip\n");
        kobject_uevent_env(&dev->primary->kdev->kobj, KOBJ_CHANGE,
                   reset_event);

        /*
         * In most cases it's guaranteed that we get here with an RPM
         * reference held, for example because there is a pending GPU
         * request that won't finish until the reset is done. This
         * isn't the case at least when we get here by doing a
         * simulated reset via debugs, so get an RPM reference.
         */
        intel_runtime_pm_get(dev_priv);

        intel_prepare_reset(dev);

        /*
         * All state reset _must_ be completed before we update the
         * reset counter, for otherwise waiters might miss the reset
         * pending state and not properly drop locks, resulting in
         * deadlocks with the reset work.
         */
        ret = i915_reset(dev); <---- THIS IS RETURNING NON-ZERO

        intel_finish_reset(dev);

        intel_runtime_pm_put(dev_priv);

        if (ret == 0) {
            /*
             * After all the gem state is reset, increment the reset
             * counter and wake up everyone waiting for the reset to
             * complete.
             *
             * Since unlock operations are a one-sided barrier only,
             * we need to insert a barrier here to order any seqno
             * updates before
             * the counter increment.
             */
            smp_mb__before_atomic();
            atomic_inc(&dev_priv->gpu_error.reset_counter);

            kobject_uevent_env(&dev->primary->kdev->kobj,
                       KOBJ_CHANGE, reset_done_event);
        } else {
            atomic_or(I915_WEDGED, &error->reset_counter); <---- WHICH GETS US HERE
        }

        /*
         * Note: The wake_up also serves as a memory barrier so that
         * waiters see the update value of the reset counter atomic_t.
         */
        i915_error_wake_up(dev_priv, true);
    }
}

[mattmacy]

/**
 * i915_reset - reset chip after a hang
 * @dev: drm device to reset
 *
 * Reset the chip.  Useful if a hang is detected. Returns zero on successful
 * reset or otherwise an error code.
 *
 * Procedure is fairly simple:
 *   - reset the chip using the reset reg
 *   - re-init context state
 *   - re-init hardware status page
 *   - re-init ring buffer
 *   - re-init interrupt state
 *   - re-init display
 */
int i915_reset(struct drm_device *dev)
{
    struct drm_i915_private *dev_priv = dev->dev_private;
    bool simulated;
    int ret;

    intel_reset_gt_powersave(dev);

    mutex_lock(&dev->struct_mutex);

    i915_gem_reset(dev);

    simulated = dev_priv->gpu_error.stop_rings != 0;

    ret = intel_gpu_reset(dev); <---- EITHER THIS FAILS

    /* Also reset the gpu hangman. */
    if (simulated) {
        DRM_INFO("Simulated gpu hang, resetting stop_rings\n");
        dev_priv->gpu_error.stop_rings = 0;
        if (ret == -ENODEV) {
            DRM_INFO("Reset not implemented, but ignoring "
                 "error for simulated gpu hangs\n");
            ret = 0;
        }
    }

    if (i915_stop_ring_allow_warn(dev_priv))
        pr_notice("drm/i915: Resetting chip after gpu hang\n");

    if (ret) {
        DRM_ERROR("Failed to reset chip: %i\n", ret);
        mutex_unlock(&dev->struct_mutex);
        return ret;
    }

    intel_overlay_reset(dev_priv);

    /* Ok, now get things going again... */

    /*
     * Everything depends on having the GTT running, so we need to start
     * there.  Fortunately we don't need to do this unless we reset the
     * chip at a PCI level.
     *
     * Next we need to restore the context, but we don't use those
     * yet either...
     *
     * Ring buffer needs to be re-initialized in the KMS case, or if X
     * was running at the time of the reset (i.e. we weren't VT
     * switched away).
     */

    /* Used to prevent gem_check_wedged returning -EAGAIN during gpu reset */
    dev_priv->gpu_error.reload_in_reset = true;

    ret = i915_gem_init_hw(dev); <---- OR THIS FAILS

    dev_priv->gpu_error.reload_in_reset = false;

    mutex_unlock(&dev->struct_mutex);
    if (ret) {
        DRM_ERROR("Failed hw init on reset %d\n", ret);
        return ret;
    }

[mattmacy]

Maybe try and instrument this to make sure we're actually waiting for 500ms here?

static int gen6_do_reset(struct drm_device *dev)
{
    struct drm_i915_private *dev_priv = dev->dev_private;
    int ret;

    /* Reset the chip */

    /* GEN6_GDRST is not in the gt power well, no need to check
     * for fifo space for the write or forcewake the chip for
     * the read
     */
    __raw_i915_write32(dev_priv, GEN6_GDRST, GEN6_GRDOM_FULL);

    /* Spin waiting for the device to ack the reset request */
    ret = wait_for((__raw_i915_read32(dev_priv, GEN6_GDRST) & GEN6_GRDOM_FULL) == 0, 500); <----- THIS MAY BE DODGY

    intel_uncore_forcewake_reset(dev, true);

    return ret;
}

[johalun]

I get "reset in progress". What API can I use to time a function in a kernel driver?

[mattmacy]

from sys/time.h:

/*
 * Functions for looking at our clock: [get]{bin,nano,micro}[up]time()
 *
 * Functions without the "get" prefix returns the best timestamp
 * we can produce in the given format.
 *
 * "bin"   == struct bintime  == seconds + 64 bit fraction of seconds.
 * "nano"  == struct timespec == seconds + nanoseconds.
 * "micro" == struct timeval  == seconds + microseconds.
 *
 * Functions containing "up" returns time relative to boot and
 * should be used for calculating time intervals.
 *
 * Functions without "up" returns UTC time.
 *
 * Functions with the "get" prefix returns a less precise result
 * much faster than the functions without "get" prefix and should
 * be used where a precision of 1/hz seconds is acceptable or where
 * performance is priority. (NB: "precision", _not_ "resolution" !)
 */

void    binuptime(struct bintime *bt);
void    nanouptime(struct timespec *tsp);
void    microuptime(struct timeval *tvp);

This is purely for instrumentation so I think the added overhead of the extra resolution is ok.
"microuptime" sounds like the way to go.

[johalun]

I don't get any output at all from gen6_do_reset()...

[johalun]

Getting -5 from intel_gpu_reset(). Btw, isn't cherryview gen8? Have to stop now but can keep digging tomorrow.

[mattmacy]

Try sticking a BACKTRACE() in each of the reset functions to see which is getting called.

[mattmacy]

gen6_do_reset is called from gen8_do_reset

[mattmacy]

I'm on #freebsd-xorg on EFnet much of the time. Easier to discuss in real-time.

[johalun]

reset request timeouts so gen6_do_reset never gets called..

[drm:cherryview_enable_rps] setting GPU freq to 400 MHz (40)
[drm] stuck on render ring
[drm] stuck on blitter ring
[drm] stuck on bsd ring
[drm] stuck on video enhancement ring
[drm] GPU HANG: ecode 8:0:0x00201001, reason: Ring hung, action: reset
[drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace.
[drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel
[drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue.
[drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it.
[drm] GPU crash dump saved to /sys/class/drm/card0/error
[drm:i915_reset_and_wakeup] resetting chip
[drm:0xffffffff82a2b4bds] *ERROR* gpu hanging too fast, banning!
intel_gpu_reset() at intel_gpu_reset+0x1e/frame 0xfffffe004996c790
i915_reset() at i915_reset+0x6d/frame 0xfffffe004996c800
i915_reset_and_wakeup() at i915_reset_and_wakeup+0x162/frame 0xfffffe004996c890
i915_handle_error() at i915_handle_error+0x11e/frame 0xfffffe004996c990
i915_hangcheck_elapsed() at i915_hangcheck_elapsed+0x49f/frame 0xfffffe004996ca60
gen8_do_reset() at gen8_do_reset+0x1e/frame 0xfffffe004996c740
intel_gpu_reset() at intel_gpu_reset+0x7a/frame 0xfffffe004996c790
i915_reset() at i915_reset+0x6d/frame 0xfffffe004996c800
i915_reset_and_wakeup() at i915_reset_and_wakeup+0x162/frame 0xfffffe004996c890
i915_handle_error() at i915_handle_error+0x11e/frame 0xfffffe004996c990
[drm:0xffffffff82ba00ccs] *ERROR* blitter ring: reset request timeout
i915_reset: intel_gpu_reset returns -5
drm/i915: Resetting chip after gpu hang
[drm:0xffffffff82a1511ds] *ERROR* Failed to reset chip: -5

[mattmacy]

Where though?

[johalun]

Not sure what this means but at the second ring reset fails.

[drm:i915_reset_and_wakeup] resetting chip
[drm:0xffffffff82a2b4bds] ERROR gpu hanging too fast, banning!
intel_gpu_reset() at intel_gpu_reset+0x1e/frame 0xfffffe0049976790
i915_reset() at i915_reset+0x6d/frame 0xfffffe0049976800
i915_reset_and_wakeup() at i915_reset_and_wakeup+0x162/frame 0xfffffe0049976890
i915_handle_error() at i915_handle_error+0x11e/frame 0xfffffe0049976990
i915_hangcheck_elapsed() at i915_hangcheck_elapsed+0x49f/frame 0xfffffe0049976a60
gen8_do_reset() at gen8_do_reset+0x21/frame 0xfffffe0049976740
intel_gpu_reset() at intel_gpu_reset+0x7a/frame 0xfffffe0049976790
i915_reset() at i915_reset+0x6d/frame 0xfffffe0049976800
i915_reset_and_wakeup() at i915_reset_and_wakeup+0x162/frame 0xfffffe0049976890
i915_handle_error() at i915_handle_error+0x11e/frame 0xfffffe0049976990
gen8_do_reset: for each ring i=0. dev=0xfffffe00016b2000. engine=0xfffffe00016b38d8.
wait_for_register() at wait_for_register+0x23/frame 0xfffffe00499766b0
gen8_do_reset() at gen8_do_reset+0x132/frame 0xfffffe0049976740
intel_gpu_reset() at intel_gpu_reset+0x7a/frame 0xfffffe0049976790
i915_reset() at i915_reset+0x6d/frame 0xfffffe0049976800
i915_reset_and_wakeup() at i915_reset_and_wakeup+0x162/frame 0xfffffe0049976890
gen8_do_reset: for each ring i=1. dev=0xfffffe00016b2000. engine=0xfffffe00016b4b78.
wait_for_register() at wait_for_register+0x23/frame 0xfffffe00499766b0
gen8_do_reset() at gen8_do_reset+0x132/frame 0xfffffe0049976740
intel_gpu_reset() at intel_gpu_reset+0x7a/frame 0xfffffe0049976790
i915_reset() at i915_reset+0x6d/frame 0xfffffe0049976800
i915_reset_and_wakeup() at i915_reset_and_wakeup+0x162/frame 0xfffffe0049976890
[drm:0xffffffff82ba0104s] ERROR blitter ring: reset request timeout
gen8_do_reset: for each ring do reset (not ready) i=0. dev=0xfffffe00016b2000. engine=0xfffffe00016b38d8.
gen8_do_reset: for each ring do reset (not ready) i=1. dev=0xfffffe00016b2000. engine=0xfffffe00016b4b78.
gen8_do_reset: for each ring do reset (not ready) i=2. dev=0xfffffe00016b2000. engine=0xfffffe00016b5e18.
gen8_do_reset: for each ring do reset (not ready) i=4. dev=0xfffffe00016b2000. engine=0xfffffe00016b8358.
i915_reset: intel_gpu_reset returns -5
drm/i915: Resetting chip after gpu hang
[drm:0xffffffff82a1511ds] ERROR Failed to reset chip: -5

[johalun]

I'm on IRC btw.

[mattmacy]

nick? I'm on #freebsd-xorg on Efnet.

[mattmacy]

Works for me!

[mattmacy]

Please file a separate issue for any further problems.