Comment 0 for bug 1877394

This is what the crash looks like:
    BUG: unable to handle page fault for address: 0000000000003448
    RIP: 0010:gen8_emit_flush_render+0x163/0x190
    Call Trace:
     execlists_request_alloc+0x25/0x40
     __i915_request_create+0x1f4/0x2c0
     i915_request_create+0x71/0xc0
     i915_gem_do_execbuffer+0xb98/0x1a80
     ? preempt_count_add+0x68/0xa0
     ? _raw_spin_lock+0x13/0x30
     ? _raw_spin_unlock+0x16/0x30
     i915_gem_execbuffer2_ioctl+0x1de/0x3c0
     ? i915_gem_busy_ioctl+0x7f/0x1d0
     ? i915_gem_execbuffer_ioctl+0x2d0/0x2d0
     drm_ioctl_kernel+0xb2/0x100
     drm_ioctl+0x209/0x360
     ? i915_gem_execbuffer_ioctl+0x2d0/0x2d0
     ksys_ioctl+0x87/0xc0
     __x64_sys_ioctl+0x16/0x20
     do_syscall_64+0x4e/0x150
     entry_SYSCALL_64_after_hwframe+0x44/0xa9

This bug was fixed by "UBUNTU: SAUCE: drm/i915: Synchronize active and retire callbacks" but there is an upstream fix for it, "drm/i915/gt: Make intel_ring_unpin() safe for concurrent pint". Let's replace the sauce patch with the upstream fix.