Bug#852749: Bug still present, patches work on modules in kernel package 4.9.0-9

2021-04-30 Thread Salvatore Bonaccorso
Control: tags -1 + moreinfo

On Fri, May 10, 2019 at 10:10:05PM -0400, AJ Milne wrote:
> Realizing this thread is now a few years old: I had the same error
> (transfer buffer not dma capable, coming from usb_hcd_map_urb_for_dma)
> using a Line6 Pod XT through the snd-usb-pod kernel module (uses
> snd-usb-line6), the machine running stock kernel package 4.9.0-9.
> 
> Inspecting the source, it appears the message buffers are still coming from
> the stack as of this kernel package. Applying Ben Hutchings' patches as of
> message:
> 
> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=852749#10
> 
> ... these moving the message buffers off the stack and adding error checks
> to the toneport set, doing so corrected the problem. Device is now
> functional; thanks much.

Can you check with more recent kernels if the issue still persist
without the patches from Ben? (I.e. has the issue in meanwhile been
fixed upstream directly)?

Regards,
Salvatore



Bug#852749: Bug still present, patches work on modules in kernel package 4.9.0-9

2019-05-10 Thread AJ Milne
Realizing this thread is now a few years old: I had the same error
(transfer buffer not dma capable, coming from usb_hcd_map_urb_for_dma)
using a Line6 Pod XT through the snd-usb-pod kernel module (uses
snd-usb-line6), the machine running stock kernel package 4.9.0-9.

Inspecting the source, it appears the message buffers are still coming from
the stack as of this kernel package. Applying Ben Hutchings' patches as of
message:

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=852749#10

... these moving the message buffers off the stack and adding error checks
to the toneport set, doing so corrected the problem. Device is now
functional; thanks much.

AJM