Messages in this thread | | | Date | Sun, 7 Aug 2011 10:20:30 -0700 | Subject | Re: avr32: handle_signal() bug? | From | Håvard Skinnemoen <> |
| |
Hi Matt,
On Wed, Aug 3, 2011 at 2:04 AM, Matt Fleming <matt@console-pimps.org> wrote: > That doesn't look correct to me. Now, if we were unsuccessful in setting > up a signal frame, say, ret == -EFAULT, do we really want to block the > signal or any of the signals in the handler mask?
I'm assuming this is a rhetorical question :-)
> Is there some intricacy of the avr32 architecture that I'm missing here? > It looks to me like this code was copied from the arm implementation > from years ago before commit a6c61e9dfdd0 ("[ARM] 3168/1: Update ARM > signal delivery and masking").
I don't think there are any avr32-specific intricacies to consider here, and ARM was indeed one of the architectures I looked at when writing the signal handling code, so I probably picked up that bug from there.
> How about this?
Looks good to me. I'm not sure how to test it though...I can try to build a kernel, run it on my board and see if it explodes, but I suspect this bug is a lot more subtle than that.
Havard
| |