diff mbox

[RFC] alsactl: Do not run udev rule before datadir is mounted

Message ID 539AC54C.60504@canonical.com (mailing list archive)
State New, archived
Headers show

Commit Message

David Henningsson June 13, 2014, 9:33 a.m. UTC 
On 2014-06-13 11:20, Takashi Iwai wrote:

>> +++ b/alsactl/90-alsa-restore.rules.in
>> @@ -1,4 +1,4 @@
>> -ACTION=="add", SUBSYSTEM=="sound", KERNEL=="controlC*", KERNELS!="card*", GOTO="alsa_restore_go"
>> +ACTION=="add", SUBSYSTEM=="sound", KERNEL=="controlC*", KERNELS!="card*", TEST=="@sbindir@", TEST=="@datadir@", GOTO="alsa_restore_go"
>>   GOTO="alsa_restore_end"
>
> IMO, it's better to use @mydatadir@ instead of @datadir@ here and...

Oops, sorry about that. Thanks for the review!

Attaching a fixup patch.

Comments

Jaroslav Kysela June 13, 2014, 10:08 a.m. UTC | #1 
Hello all,

  ALSA 1.0.28 packages are available for testing at

ftp://ftp.alsa-project.org/pub/testing/

  Please, report any issues. I expect to announce the release on Monday.

					Jaroslav

-----
Jaroslav Kysela <perex@perex.cz>
Linux Kernel Sound Maintainer
ALSA Project; Red Hat, Inc.
David Henningsson June 13, 2014, 10:52 a.m. UTC | #2 
On 2014-06-13 12:08, Jaroslav Kysela wrote:
> Hello all,
>
>    ALSA 1.0.28 packages are available for testing at
>
> ftp://ftp.alsa-project.org/pub/testing/
>
>    Please, report any issues. I expect to announce the release on Monday.

Thanks!

I'm testing build of alsa-lib now and I have two questions:

1) the final so version seems to be libasound.so.2.0.0 for both 1.0.27 
and 1.0.28. I'm not sure how important it is, but shouldn't this be 
increased to 2.0.1 or 2.1.0 to reflect code changes?

2) This probably is not a new issue, but I was wondering about all the 
extra files created in configure.ac, e g ctl_symbols_list.c. Are they 
ever cleaned by some "clean", "maintainerclean" or similar rule?
Alexander Patrakov June 13, 2014, 10:59 a.m. UTC | #3 
13.06.2014 16:08, Jaroslav Kysela ?????:
> Hello all,
>
>    ALSA 1.0.28 packages are available for testing at
>
> ftp://ftp.alsa-project.org/pub/testing/
>
>    Please, report any issues. I expect to announce the release on Monday.

/usr/share/alsa/cards/Loopback.conf talks about "Intel HD audio 
(ICH6/ICH7)". This looks suspicious.
Jaroslav Kysela June 13, 2014, 5:17 p.m. UTC | #4 
Date 13.6.2014 12:52, David Henningsson wrote:
> 
> 
> On 2014-06-13 12:08, Jaroslav Kysela wrote:
>> Hello all,
>>
>>    ALSA 1.0.28 packages are available for testing at
>>
>> ftp://ftp.alsa-project.org/pub/testing/
>>
>>    Please, report any issues. I expect to announce the release on Monday.
> 
> Thanks!
> 
> I'm testing build of alsa-lib now and I have two questions:
> 
> 1) the final so version seems to be libasound.so.2.0.0 for both 1.0.27 
> and 1.0.28. I'm not sure how important it is, but shouldn't this be 
> increased to 2.0.1 or 2.1.0 to reflect code changes?

I think that we had some discussion about this and we use different
system to identify new libraries:

1) there is snd_asoundlib_version() function
2) the packaging system identifies the exact used library version
3) we use deprecated linker warning for obsolete symbols
4) only "current" libtool version number will be changed when
   the library is not backwards compatible, otherwise all symbols must
   be present and compatible

Basically, we haven't found the current:revision:age numbering useable
for our rules and as the libtool documentation states, it should not
reflect the release number at all.

Even if the age is increased, nobody will have/known a correlation
between the package release version and the library version. It means -
we would update a number for nothing in this case.

Eventually (contradicting to the paragraphs above) the age might reflect
the release number in form like:

1.0.28.3  ->  2:0:283

Future alsa-lib numbering:

2.33.5 -> 3:0:335 or 2:1:335 (if the library is backward compatible)

And this makes the question, if the version of packages should be
increased to 2.0.0 in next release to lower the last number :-)

> 2) This probably is not a new issue, but I was wondering about all the 
> extra files created in configure.ac, e g ctl_symbols_list.c. Are they 
> ever cleaned by some "clean", "maintainerclean" or similar rule?

Nobody has required this. Patches are welcome. I'm using the 'git
archive' to obtain the sources without these extra files for the
packaging purposes and 'make dist'.

				Jaroslav
Jaroslav Kysela June 13, 2014, 5:19 p.m. UTC | #5 
Date 13.6.2014 12:59, Alexander E. Patrakov wrote:
> 13.06.2014 16:08, Jaroslav Kysela ?????:
>> Hello all,
>>
>>    ALSA 1.0.28 packages are available for testing at
>>
>> ftp://ftp.alsa-project.org/pub/testing/
>>
>>    Please, report any issues. I expect to announce the release on Monday.
> 
> /usr/share/alsa/cards/Loopback.conf talks about "Intel HD audio 
> (ICH6/ICH7)". This looks suspicious.
> 

Thanks. Fixed.

                    Jaroslav
Sergey June 14, 2014, 12:11 p.m. UTC | #6 
Fri, 13 Jun 2014, Jaroslav Kysela wrote:

>   ALSA 1.0.28 packages are available for testing at
> 
> ftp://ftp.alsa-project.org/pub/testing/
> 
>   Please, report any issues. I expect to announce the release on Monday.

I get "Assertion failed" with plug for float data.
No patch yet. I'm still trying to understand how "plug" works.
The bug was in earlier alsa versions too. Was almost fixed by 2da7b0c2
(discussion http://thread.gmane.org/gmane.linux.alsa.devel/119358)

I first found it with a complex ladspa upmix config, like:
  http://forums.gentoo.org/viewtopic-p-4528619.html#4528619
An easy way to see the assertion:
pcm.!default {
    type plug
    slave.pcm { type null }
    ttable.0.0 1
}
# aplay -fFLOAT_LE /dev/zero
Playing raw data '/dev/zero' : Float 32 bit Little Endian, Rate 8000 Hz, Mono
aplay: pcm_plug.c:388: snd_pcm_plug_change_channels: Assertion `snd_pcm_format_linear(slv->format)' failed.
Aborted by signal Aborted...
diff mbox

Patch

From 7402d0421df6782dd7af9a30f3bc1223057cd9f3 Mon Sep 17 00:00:00 2001
From: David Henningsson <david.henningsson@canonical.com>
Date: Fri, 13 Jun 2014 11:31:44 +0200
Subject: [PATCH] alsactl: Fixup test for mydatadir/sbindir

Use the "mydatadir" name consistently.

Signed-off-by: David Henningsson <david.henningsson@canonical.com>
---
 alsactl/90-alsa-restore.rules.in | 2 +-
 alsactl/Makefile.am              | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/alsactl/90-alsa-restore.rules.in b/alsactl/90-alsa-restore.rules.in
index 80fc043..c0c1b23 100644
--- a/alsactl/90-alsa-restore.rules.in
+++ b/alsactl/90-alsa-restore.rules.in
@@ -1,4 +1,4 @@ 
-ACTION=="add", SUBSYSTEM=="sound", KERNEL=="controlC*", KERNELS!="card*", TEST=="@sbindir@", TEST=="@datadir@", GOTO="alsa_restore_go"
+ACTION=="add", SUBSYSTEM=="sound", KERNEL=="controlC*", KERNELS!="card*", TEST=="@sbindir@", TEST=="@mydatadir@", GOTO="alsa_restore_go"
 GOTO="alsa_restore_end"
 
 LABEL="alsa_restore_go"
diff --git a/alsactl/Makefile.am b/alsactl/Makefile.am
index afbedb3..47f06e9 100644
--- a/alsactl/Makefile.am
+++ b/alsactl/Makefile.am
@@ -44,7 +44,7 @@  endif
 
 edit = \
 	$(SED) -r -e 's,@sbindir\@,$(sbindir),g' \
-		  -e 's,@datadir\@,$(mydatadir),g' \
+		  -e 's,@mydatadir\@,$(mydatadir),g' \
 		  -e 's,@daemonswitch\@,$(ALSACTL_DAEMONSWITCH),g' \
 							< $< > $@ || rm $@
 
-- 
1.9.1