Anders Larsen schrieb:
> Hi,
>
> On Wed, 28 Mar 2007 16:05:56 +0200, Denis Oliver Kropp wrote:
>> If 0xbe826a2c is the intruction pointer you should be able to find
>> out which function it is in.
>
> On ARM this address is in the stack segment :-(
It would be good to know what the address indicates, if it's
a bad instruction or data pointer, but I think it's the former.
The sigaction(2) man page doesn't say that. I guess we have
to read kernel code.
>>>>> (!) FUSION_PROPERTY_LEASE --> Connection timed out
>> Most likely due to the clock skew, or a dead lock. Do you know if there
>> was a feelable delay between this and the previous message?
>
> The patch below seems to cure the "Connection timed out" but
> otherwise did not change the manifestation of the problem :-(
>
> Cheers
> Anders
>
> --- a/linux/drivers/char/fusion/entries.c
> +++ b/linux/drivers/char/fusion/entries.c
> @@ -403,12 +403,12 @@ fusion_entry_wait( FusionEntry *entry, l
>
> entry->waiters--;
>
> - if (timeout && !*timeout)
> - return -ETIMEDOUT;
> -
> if (signal_pending(current))
> return -EINTR;
>
> + if (timeout && !*timeout)
> + return -ETIMEDOUT;
> +
> ret = fusion_entry_lock( entries, id, false, &entry2 );
> switch (ret) {
> case -EINVAL:
Oh, great, thanks for fixing that error. I fixed a similar case a few
days ago.
This fixes the clock skew problem in this case, but what if there's no
signal?
--
Best regards,
Denis Oliver Kropp
.------------------------------------------.
| DirectFB - Hardware accelerated graphics |
| http://www.directfb.org/ |
"------------------------------------------"
_______________________________________________
directfb-dev mailing list
[email protected]
http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-dev
