Re: Sleeping from invalid context in udlfb

On Thu, 2 Aug 2018, David Airlie wrote:

I'm pretty sure udlkms handles this already.

Dave.

But it crashes on unplug :-)

Mikulas

On Wed, Aug 1, 2018 at 11:34 PM, Mikulas Patocka [off-list ref] wrote:


      On Wed, 1 Aug 2018, Geert Uytterhoeven wrote:

      > Hi Mikulas,
      >
      > On Wed, Aug 1, 2018 at 12:59 PM Mikulas Patocka [off-list ref] wrote:
      > > On Wed, 1 Aug 2018, Geert Uytterhoeven wrote:
      > > > On Tue, Jul 31, 2018 at 5:23 PM Mikulas Patocka [off-list ref] wrote:
      > > > > BTW when using the udlfb driver as a console, I've got this warning.
      > > > > vt_console_print takes a spinlock and then calls the framebuffer driver
      > > > > that sleeps.
      > > > >
      > > > > The question is - whose fault is this? Could the console code somehow be
      > > > > told to print characters without holding a spinlock? Or does it mean that
      > > > > framebuffer drivers can't sleep?
      > > > >
      > > > > udlfb communicates through USB, so the sleeping is inevitable.
      > > > >
      > > > > Mikulas
      > > > >
      > > > >
      > > > > BUG: sleeping function called from invalid context at mm/slab.h:421
      > > > > in_atomic(): 1, irqs_disabled(): 0, pid: 430, name: kworker/2:3
      > > > > 6 locks held by kworker/2:3/430:
      > > > >  #0: 000000001301127e ( (wq_completion)"events"){....} , at: process_one_work+0x17c/0x3a8
      > > > >  #1: 00000000beacc951 ( (work_completion)(&(&dlfb->init_framebuffer_work)->work)){....} , at: process_one_work+0x17c/0x3a8
      > > > >  #2: 00000000a402f826 ( registration_lock){....} , at: register_framebuffer+0x28/0x2c0 [fb]
      > > > >  #3: 0000000021cbe902 ( console_lock){....} , at: register_framebuffer+0x258/0x2c0 [fb]
      > > > >  #4: 0000000096d51735 ( console_owner){....} , at: console_unlock+0x174/0x500
      > > > >  #5: 00000000faa7f206 ( printing_lock){....} , at: vt_console_print+0x60/0x3a0
      > > > > Preemption disabled at: [<ffffff8008403130>] vt_console_print+0x60/0x3a0
      > > > > CPU: 2 PID: 430 Comm: kworker/2:3 Not tainted 4.17.10-debug #3
      > > > > Hardware name: Marvell Armada 8040 MacchiatoBin/Armada 8040 MacchiatoBin, BIOS EDK II Jul 30 2018
      > > > > Workqueue: events dlfb_init_framebuffer_work [udlfb]
      > > > > Call trace:
      > > > >  dump_backtrace+0x0/0x150
      > > > >  show_stack+0x14/0x20
      > > > >  dump_stack+0x8c/0xac
      > > > >  ___might_sleep+0x140/0x170
      > > > >  __might_sleep+0x50/0x88
      > > > >  __kmalloc+0x1b0/0x270
      > > > >  xhci_urb_enqueue+0xa8/0x460 [xhci_hcd]
      > > > >  usb_hcd_submit_urb+0xc0/0x998 [usbcore]
      > > > >  usb_submit_urb+0x1e0/0x518 [usbcore]
      > > > >  dlfb_submit_urb+0x38/0x98 [udlfb]
      > > > >  dlfb_handle_damage.isra.4+0x1e0/0x210 [udlfb]
      > > > >  dlfb_ops_imageblit+0x28/0x38 [udlfb]
      > > > >  soft_cursor+0x15c/0x1d8 [fb]
      > > > >  bit_cursor+0x324/0x510 [fb]
      > > > >  fbcon_cursor+0x144/0x1a0 [fb]
      > > > >  hide_cursor+0x38/0xa0
      > > > >  vt_console_print+0x334/0x3a0
      > > > >  console_unlock+0x274/0x500
      > > > >  register_framebuffer+0x22c/0x2c0 [fb]
      > > > >  dlfb_init_framebuffer_work+0x1ec/0x2fc [udlfb]
      > > > >  process_one_work+0x1e8/0x3a8
      > > > >  worker_thread+0x44/0x418
      > > > >  kthread+0x11c/0x120
      > > > >  ret_from_fork+0x10/0x18
      > > >
      > > > This is sort of expected: you cannot do USB transfers from printk().
      > > >
      > > > Gr{oetje,eeting}s,
      > > >
      > > >                         Geert
      > >
      > > So, should there be a framebuffer flag that prevents the console from
      > > binding to it?
      > >
      > > If I start the kernel with "console=ttyS0,115200", it doesn't try to bind
      > > to the udlfb driver, but if I start it without this flag, it does and
      > > crashes :-(
      >
      > Your frame buffer driver should offload tasks that may sleep to e.g. a
      > workqueue.
      >
      > Gr{oetje,eeting}s,
      >
      >                         Geert

I can try to do this - but - taking a spinlock and copying 8MB framebuffer
would damage scheduling latency even for PCI framebuffer drivers.

Mikulas