- 1. spotted in the kaio patch (score: 1)
- Author: Andrew Morton <andrewm@xxxxxxxxxx>
- Date: Sun, 18 Mar 2001 23:57:37 +0000
- I was surfing this patch... do_kaio_generic_read() does a page_cache_alloc() with the pagecache_lock held - can deadlock. In kaio_set_slave_name(), do this: current->mm->arg_start = current->mm->arg_
- /archives/kaio/2001-03/msg00000.html (6,365 bytes)
- 2. Re: spotted in the kaio patch (score: 1)
- Author: Rajagopal Ananthanarayanan <ananth@xxxxxxx>
- Date: Mon, 19 Mar 2001 10:23:05 -0800
- I'll look into this ... Ah, so that's the trick, thanks. I had always wanted to fix this, but never got around to it. Interestingly top reports the correct names even without this fix. Just curious:
- /archives/kaio/2001-03/msg00001.html (7,878 bytes)
- 3. Re: spotted in the kaio patch (score: 1)
- Author: Andrew Morton <andrewm@xxxxxxxxxx>
- Date: Mon, 19 Mar 2001 22:39:25 +0000
- mm.. I assume `ps' sees an empty argv and then falls back to It seems that `top' uses ->comm[] all the time, unless you give it the 'c' option. It makes me want to read the raw I/O patch :) Question,
- /archives/kaio/2001-03/msg00003.html (8,358 bytes)
- 4. Re: spotted in the kaio patch (score: 1)
- Author: Rajagopal Ananthanarayanan <ananth@xxxxxxx>
- Date: Mon, 19 Mar 2001 15:41:10 -0800
- Thank you for finding bugs in it ;-) Well, not really. kiobufs are limited to "normal" file I/O. People who care about copying will use O_DIRECT or avoid it completely by using RAW devices ... most d
- /archives/kaio/2001-03/msg00004.html (8,916 bytes)
- 5. spotted in the kaio patch (score: 1)
- Author: Andrew Morton <andrewm@xxxxxxxxxx>
- Date: Sun, 18 Mar 2001 23:57:37 +0000
- I was surfing this patch... do_kaio_generic_read() does a page_cache_alloc() with the pagecache_lock held - can deadlock. In kaio_set_slave_name(), do this: current->mm->arg_start = current->mm->arg_
- /archives/kaio/2001-03/msg00006.html (6,365 bytes)
- 6. Re: spotted in the kaio patch (score: 1)
- Author: Rajagopal Ananthanarayanan <ananth@xxxxxxx>
- Date: Mon, 19 Mar 2001 10:23:05 -0800
- I'll look into this ... Ah, so that's the trick, thanks. I had always wanted to fix this, but never got around to it. Interestingly top reports the correct names even without this fix. Just curious:
- /archives/kaio/2001-03/msg00007.html (7,903 bytes)
- 7. Re: spotted in the kaio patch (score: 1)
- Author: Andrew Morton <andrewm@xxxxxxxxxx>
- Date: Mon, 19 Mar 2001 22:39:25 +0000
- mm.. I assume `ps' sees an empty argv and then falls back to It seems that `top' uses ->comm[] all the time, unless you give it the 'c' option. It makes me want to read the raw I/O patch :) Question,
- /archives/kaio/2001-03/msg00009.html (8,408 bytes)
- 8. Re: spotted in the kaio patch (score: 1)
- Author: Rajagopal Ananthanarayanan <ananth@xxxxxxx>
- Date: Mon, 19 Mar 2001 15:41:10 -0800
- Thank you for finding bugs in it ;-) Well, not really. kiobufs are limited to "normal" file I/O. People who care about copying will use O_DIRECT or avoid it completely by using RAW devices ... most d
- /archives/kaio/2001-03/msg00010.html (8,991 bytes)
This search system is powered by
Namazu