Skip to contentSkip to conversation

: API compatibility with pselect(2)

ALSA - lib 0000898 [email protected] 1 participant 1 message Jul 8, 2005

lore
Share & tools
lore
#0 lore raw

The following issue has been CLOSED
======================================================================
<https://bugtrack.alsa-project.org/alsa-bug/view.php?id=898>
======================================================================
Reported By: phunge
Assigned To:
======================================================================

Project:                    ALSA - lib
Issue ID:                   898
Category:                   0_general
Reproducibility:            always
Severity:                   feature
Priority:                   normal
Status:                     closed
Resolution:                 not an issue
Fixed in Version:           
======================================================================
Date Submitted:             02-09-2005 16:11 CET
Last Modified:              07-08-2005 10:35 CEST

======================================================================
Summary: API compatibility with pselect(2)
Description:
I have a program which uses the ALSA sequencer API, and also uses setitimer() to trigger some regular processing.

I had a bug where I cannot open the sequencer when running under valgrind
- the
program output "Unknown SEQ default" when calling snd_seq_open. My theory is, this is caused by a SIGALRM being delivered while alsa-lib is in read() or write() or some other syscall, and none of alsa-lib handles EINTR correctly.

I fixed this by blocking SIGALRM except for when calling poll(). However, using poll() there's no way to atomically unblock a signal and enter poll() - that's what pselect is for. However, the snd_*_poll_descriptors-style API makes it hard to use select/pselect, so I'm forced to implement it in a way that's not 100%
free of race conditions.

My request: please add a way to use select/pselect with ALSA.
======================================================================

----------------------------------------------------------------------
phunge - 02-14-05 11:04
----------------------------------------------------------------------
I didn't think of that - that will work fine. This bug is a non-issue.

Thanks for the info.

----------------------------------------------------------------------
tiwai - 02-14-05 16:09
----------------------------------------------------------------------
OK, closing now.

Issue History
Date Modified Username Field Change
======================================================================

02-09-05 16:11 phunge         New Issue                                    
02-09-05 17:12 tiwai          Note Added: 0003557                          
02-14-05 11:04 phunge         Note Added: 0003588                          
02-14-05 16:09 tiwai          Status                   new => resolved     
02-14-05 16:09 tiwai          Resolution               open => not an issue
02-14-05 16:09 tiwai          Assigned To               => tiwai           
02-14-05 16:09 tiwai          Note Added: 0003591                          
02-14-05 16:09 tiwai          Assigned To              tiwai =>            
07-08-05 10:35 perex          Status                   resolved => closed  

======================================================================

-------------------------------------------------------
This SF.Net email is sponsored by the 'Do More With Dual!' webinar happening July 14 at 8am PDT/11am EDT. We invite you to explore the latest in dual core and dual graphics technology at this free one hour event hosted by HP, AMD, and NVIDIA. To register visit http://www.hp.com/go/dualwebinar