Matthias Urlichs wrote:
> Hi,
>
> Hans Verkuil:
>
>>> One thing to note is that with hyperthreading enabled, the system
>>> keeps running fine after a timeout, but with hyperthreading disabled
>>> the system freezes and reboots. (untested disabling on the lastest
>>> build, this was 2.6.19.1 and 0.10.3)
>>>
>> Really? I'd better put that on my lists of things to check. Lately I've
>> been using multi-core CPUs almost exclusively, so I probably wouldn't
>> notice a bug like that.
>>
>>
> Maybe the problem wedges one of the CPUs, regardless of whether HT is on
> or off?
>
>
A) Well, it may do so, but in HT mode, it gets 'unstuck' too. I can have
multiple cards freeze up (like 6) and still have the machine running ok.
B)
Doing a bit of research into this, the last lockup I had was after a
0x1C000 byte size DMA transfer, thats pretty big compared to the usual ones
After a bit of sampling I see DMA size distribution to be mainly in the
0x2800 to 0x8000 region, a few crashes i provoked by setting the latency
down to 1, had transfers set up just beforehand that were bigger than
this: 0x8800, 0x12000, 0xE000, 0xA000
I was looking at the code to see if it would be possible to fallback to
a PIO mode transfer for larger than 0x8000 sized transfers. Would this
be possible to do easily? Or how about multiple smaller DMA transfers?
C) ivtv-irq.c: ivtv_unfinished_dma() does not call del_timer, unlike
every other routine, I assume this is because it is how it was invoked,
does this matter? I added it as a test, it doesnt fix/break anything
obvious.
D ) FYI, there is an also odd distribution of DMA sizes that I'm
intrigued by. This list is heavily trimmed either size, but shows the
pattern clearly. I'm not capturing VBI data either.
[occurrences and sizes]
2729 0x00002800
120 0x00002840
129 0x00002fc0
5497 0x00003000
101 0x00003040
111 0x000037c0
7569 0x00003800
89 0x00003840
89 0x00003fc0
7238 0x00004000
73 0x00004040
60 0x000047c0
3295 0x00004800
45 0x00004840
62 0x00004fc0
Regards,
Mark Bryars
--
Mark Bryars
Product Development Engineer
ETV Interactive Ltd
Logie Court
Stirling University Innovation Park
Stirling
Scotland, UK
FK9 4NF
T: +44 (0) 1786 455150
F: +44 (0) 1786 455179
W: www.etvinteractive.com
_______________________________________________
ivtv-devel mailing list
[email protected]
http://ivtvdriver.org/mailman/listinfo/ivtv-devel
