Re: [gpm] Segfault on button-pressed

[Richard Hughes <hughsient gmail com>]

On Tue, 2006-11-21 at 21:00 +0100, Pascal Terjan wrote:
> Hello,

Sorry for the delay, it's been a busy week.

> Today when suspending I got a crash in :
> #4  0x0804f5a4 in emit_button_pressed (monitor=0x807d168,
>     udi=0x80b5a88 "/org/freedesktop/Hal/devices/acpi_LID", details=)
>     at gpm-hal-monitor.c:193
>         button_name = 0x0
>         value = 1
>         __PRETTY_FUNCTION__ = "emit_button_pressed"
>         __FUNCTION__ = "emit_button_pressed"
> 
> So it looks like Hal informed us that the lid button was pressed but
> then could not give us the button name. I think this is a race when
> the button is pressed while hal is shutting down.

Ahh, gotcha.

> Here is a small patch that would avoid the crash.
> Maybe testing the return of gpm_hal_device_get_string would be nicer.

For 2.16, I think your patch is fine. Please could you apply this to
2-16, and if you've got a spare 5 minutes, could you also do a similar
fix for HEAD (look in gpm-button.c, emit_button_pressed).

Thanks for squashing this bug.

Richard.