NetBSD-Bugs archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: PR/57436 CVS commit: src/sys/dev/pad

The following reply was made to PR kern/57436; it has been noted by GNATS.

From: (Michael van Elst)
Subject: Re: PR/57436 CVS commit: src/sys/dev/pad
Date: Sat, 27 May 2023 05:38:50 -0000 (UTC) ("Nathanial Sloss") writes:
 > Module Name:	src
 > Committed By:	nat
 > Date:		Fri May 26 10:39:56 UTC 2023
 > Modified Files:
 > 	src/sys/dev/pad: pad.c
 > Log Message:
 > Fix logic for timing.
 This and the following change break the "high precision" timing
 that I added for netbsd-10.
 In netbsd-9 pad(4) would just block until the whole userland buffer
 was filled, making it impossible to actually 'time' I/O to what
 the audio driver produces unless you know exactly what data it will
 produce (or do single sample reads).
 With the 'high precision' timing, you would get short reads with
 data blocks as they get produced. This might be what broke bta2dpd.
 With this 'fix' pad(4) will just return end-of-file when no data is
 available, breaking almost every user.

Home | Main Index | Thread Index | Old Index