Bug#1051445: syslinux.efi crashes on isohybrid boot as cdrom

[Ralph Ronnquist]

Note that a debian package patch on 3:6.04~git20190206.bf6db5b4+dfsg1-3
is available at:

https://git.devuan.org/devuan/syslinux/raw/branch/suites/experimental/debian/patches/0020-fix-cdrom-on-uefi.patch

That's the patch used for the Devuan experimental release package.
Presumably it can be used directly a Debian pcakage upgrade.

regards,

Ralph.


signature.asc
Description: PGP signature

Bug#1051445: syslinux.efi crashes on isohybrid boot as cdrom

[Ralph Ronnquist]

Package: syslinux-efi
Version: 6.04~git20190206.bf6db5b4+dfsg1

The use case of a hybrid ISO where an EFI partition with syslinux.efi
is set up as El Torrito boot option, all duly equipped. works fine
when the media is provided as disk drive, but not when the same media
is provided as cdrom drive.

The cause seems to be in a mixup between the actual and assumed media
sector sizes; a cdrom has 2048 byte sectors while a disk has 256 byte
sectors. (In particular the partition table on the ISO has sector
indexes based on 256 byte sectors)

As an attempt to resolve that, I made teh attached patch for
efi/diskio.c. It resolves the sector size "mixup" by means of a
wrapper function for disk reading/writing that translates between
(hardcoded) "wanted" sectors of 256 byte size, and "actual"
(bio->Media->BlockSize) when different from that. The implementation
lacks some in efficiency but works well enough for boot loading.

I also attch the patch for efi/efi.h I needed in order to build the
package.

Hopefully this patching can be applied to the syslinux-efi package, to
allow for isohybrid ISO that has only isolinux/syslinux as boot setup.

Ralph.
--- syslinux-6.04~git20190206.bf6db5b4+dfsg1/efi/diskio.c   2019-02-06 
19:30:51.0 +
+++ fixed/efi/diskio.c  2023-09-08 05:11:08.646579577 +
@@ -6,6 +6,7 @@
 #include 
 #include 
 #include 
+#include 
 #include "efi.h"
 
 static inline EFI_STATUS read_blocks(EFI_BLOCK_IO *bio, uint32_t id, 
@@ -26,7 +27,7 @@
struct efi_disk_private *priv = (struct efi_disk_private 
*)disk->private;
EFI_BLOCK_IO *bio = priv->bio;
EFI_STATUS status;
-   UINTN bytes = count * disk->sector_size;
+   UINTN bytes = count * bio->Media->BlockSize;
 
if (is_write)
status = write_blocks(bio, disk->disk_number, lba, bytes, buf);
@@ -38,7 +39,70 @@
is_write ? L"write" : L"read",
status);
 
-   return count << disk->sector_shift;
+   return bytes;
+}
+
+/**
+ * Map read/write of EFI sectors of 512 bytes into read/write of iso
+ * sectors of 2048 bytes.
+ *
+ * real_lba = lba / 4; real_count = ( count + 3 ) / 4;
+ */
+static int efi_iso_rdwr_sectors(struct disk *disk, void *buf,
+   sector_t lba, size_t count, bool is_write) {
+  static int ratio = 0;
+
+  struct efi_disk_private *priv = (struct efi_disk_private *)disk->private;
+  EFI_BLOCK_IO *bio = priv->bio; 
+  size_t bytes = count * disk->sector_size;
+
+  if ( ! ratio ) {
+ratio = bio->Media->BlockSize / disk->sector_size;
+  }
+
+  sector_t real_lba = lba / ratio;
+  size_t real_count = ( count + ratio - 1 ) / ratio;
+  size_t size = bytes + bio->Media->BlockSize -disk->sector_size;
+
+  void *buffer = malloc( size );
+  
+  void *start = buffer + ( lba % ratio ) * disk->sector_size;
+
+  int err = 0;
+  
+  if ( !buffer ) {
+return EFI_OUT_OF_RESOURCES;
+  }
+  if ( is_write ) {
+// read the first block from disk if head padding is needed
+if ( buffer != start ) {
+  err = efi_rdwr_sectors( disk, buffer, real_lba, 1, 0 );
+  if ( err < 0 ) goto ending;
+}
+// load last block from disk if tail padding is needed
+if ( ( real_count > 1 ) && ( start + bytes < buffer + size ) ) {
+  void *last = buffer + size - bio->Media->BlockSize;
+  sector_t end_lba = real_lba + real_count - 1;
+  err = efi_rdwr_sectors( disk, last, end_lba, 1, 0 );
+  if ( err < 0 ) goto ending;
+}
+memcpy( start, buf, bytes );
+err = efi_rdwr_sectors( disk, buffer, real_lba, real_count, 1 );
+if ( err < 0 ) goto ending;
+// Assume the whole buffer was written
+err = bytes;
+  } else {
+// read into buffer
+err = efi_rdwr_sectors( disk, buffer, real_lba, real_count, 0 );
+if ( err < 0 ) goto ending;
+// Assume the whole buffer was read
+err = bytes;
+memcpy( buf, start, err );
+  }
+
+ ending:
+  if ( buffer ) free( buffer );
+  return err;
 }
 
 struct disk *efi_disk_init(void *private)
@@ -65,8 +129,9 @@
  */
 disk.disk_number   = bio->Media->MediaId;
 
-disk.sector_size   = bio->Media->BlockSize;
-disk.rdwr_sectors  = efi_rdwr_sectors;
+disk.sector_size   = 512; // bio->Media->BlockSize;
+disk.rdwr_sectors  = ( bio->Media->BlockSize == 512 )?
+  efi_rdwr_sectors : efi_iso_rdwr_sectors;
 disk.sector_shift  = ilog2(disk.sector_size);
 
 dprintf("sector_size=%d, disk_number=%d\n", disk.sector_size,
--- syslinux-6.04~git20190206.bf6db5b4+dfsg1/efi/efi.h  2019-02-06 
19:30:51.0 +
+++ fixed/efi/efi.h 2023-09-05 14:29:41.227060161 +
@@ -21,6 +21,7 @@
 #endif
 
 #include 
+#include 
 #include 
 #include 
 


signature.asc
Description: PGP signature

Bug#1039687: RFS: fusefile/1.0 [ITP] -- Using FUSE to combine file fragments into a single file.

[Ralph Ronnquist]

Package: sponsorship-requests
Severity: wishlist

Dear mentors,

I am looking for a sponsor for my package "fusefile":

 * Package name : fusefile
   Version  : 1.0
   Upstream contact : Ralph Ronnquist 
 * URL  : https://git.devuan.org/rrq/fusefile.git
 * License  : GPL-3.0+
 * Vcs  : [fill in URL of packaging vcs]
   Section  : admin

The source builds the following binary packages:

  fusefile - Using FUSE to combine file fragments into a single file.

To access further information about this package, please visit the following 
URL:

  https://mentors.debian.net/package/fusefile/

Alternatively, you can download the package with 'dget' using this command:

  dget -x 
https://mentors.debian.net/debian/pool/main/f/fusefile/fusefile_1.0.dsc

Changes for the initial release:

 fusefile (1.0) unstable; urgency=medium
 .
   * Version change for submission to debian

Regards,
-- 
  Ralph Ronnquist

Bug#1039062: RFS: fusefile/1.0 [ITP] -- This package provides "fusefile"; a FUSE file mount that

[Ralph Ronnquist]

Package: sponsorship-requests
Severity: wishlist

Dear mentors,

I am looking for a sponsor for my package "fusefile":

 * Package name : fusefile
   Version  : 1.0
   Upstream contact : [fill in name and email of upstream]
 * URL  : https://git.devuan.org/rrq/fusefile.git
 * License  : GPL-3.0+
 * Vcs  : [fill in URL of packaging vcs]
   Section  : admin

The source builds the following binary packages:

  fusefile - This package provides "fusefile"; a FUSE file mount that

To access further information about this package, please visit the following 
URL:

  https://mentors.debian.net/package/fusefile/

Alternatively, you can download the package with 'dget' using this command:

  dget -x 
https://mentors.debian.net/debian/pool/main/f/fusefile/fusefile_1.0.dsc

Changes for the initial release:

 fusefile (1.0) unstable; urgency=medium
 .
   * Version change for submission to debian

Regards,
-- 
  Ralph Ronnquist


signature.asc
Description: PGP signature

Bug#1029603: ...

[Ralph Ronnquist]

Thanks. As it turns out, I only need to have the pathname
/proc/self/fd/5 exist to be chmod-able, and then emacs-nox installs
without isse. Thus, the commands:
  # mkdir -p /proc/self/fd
  # touch /proc/self/fd/5
are enough to set the scene for the successful installation of
emacs-nox.

It does seem a little bit odd that that installation calls "chmod 644"
on "/proc/self/fd/5", though maybe that is less odd than me making
this pathname a normal file rather than a "proc" mount :)

regards,

Ralph.


signature.asc
Description: PGP signature

Bug#1029603: ...

[Ralph Ronnquist]

I have now tried this on qemu using kernel 6.1.0-1-amd64.
Same result.

Test sequence is:
# debootstrap --variant=minbase sid FS http://deb.debian.org/debian
# chroot FS apt-get install emacs-nox

Results in:
emacsen-common: Handling install of emacsen flavor emacs
>>Error occurred processing /usr/share/emacs/site-lisp/debian-startup.el: File 
>>error (("Doing chmod" "Operation not supported" 
>>"/usr/share/emacs/site-lisp/debian-startup.elcQwWhCN"))
ERROR: install script from emacsen-common package failed

etc.

Ralph.


signature.asc
Description: PGP signature

Bug#1029603: ...

[Ralph Ronnquist]

This is a debootstrap filesystem; I just repeated it with a new
debootstrapped sid filesystem.

Since you think this might be a kernel related issue I'll check it out
with a qemu as well.

Ralph.

Bug#1029603: emacsen-common fails to install due to attempted chmod of /proc/self/fd/4

[Ralph Ronnquist]

Package: emacsen-common
Version: 3.0.5
Severity: grave
Justification: renders package unusable
X-Debbugs-Cc: ralph.ronnqu...@gmail.com

Dear Maintainer,

   * What led up to the situation?
this problem came up when I tried installing emacs-nox, with
# apt-get install --no-install-recommends emacs-nox

* What was the outcome of this action?
Install emacsen-common for emacs
emacsen-common: Handling install of emacsen flavor emacs
>>Error occurred processing /usr/share/emacs/site-lisp/debian-startup.el: File 
>>error (("Doing chmod" "Operation not supported" 
>>"/usr/share/emacs/site-lisp/debian-startup.elcJzWI2I"))
ERROR: install script from emacsen-common package failed
dpkg: error processing package emacs-nox (--configure):
 installed emacs-nox package post-installation script subprocess returned error 
exit status 1
Processing triggers for libc-bin (2.36-8) ...
Processing triggers for man-db (2.11.2-1) ...
Processing triggers for install-info (6.8-6+b1) ...
Errors were encountered while processing:
 emacs-nox
E: Sub-process /usr/bin/dpkg returned an error code (1)

   * What outcome did you expect instead?
Install emacsen-common for emacs
emacsen-common: Handling install of emacsen flavor emacs
Processing triggers for install-info (6.8-6+b1) ...
Processing triggers for man-db (2.11.2-1) ...

Note, when I ran it under "strace -f", it shows that the ultimate
failure cause is an attempt to "chmod /proc/self/fd/4".

-- System Information:
Debian Release: bookworm/sid
merged-usr: no
Architecture: amd64 (x86_64)

Kernel: Linux 5.10.0-20-amd64 (SMP w/8 CPU threads)
Locale: LANG=en_AU.UTF-8, LC_CTYPE=C.UTF-8 (charmap=locale: Cannot set 
LC_MESSAGES to default locale: No such file or directory
locale: Cannot set LC_ALL to default locale: No such file or directory
UTF-8), LANGUAGE not set
Shell: /bin/sh linked to /bin/dash
Init: sysvinit (via /sbin/init)

Bug#1013916: missing dependency

[Ralph Ronnquist]

Package: cron-daemon-common
Version: 3.0pl1-144

This package seems to have a dependency fault in that it needs
"addgroup" as part of its postinst but doesn't declare a dependency on
the package "adduser" that provides "addgroup".

This e.g. causes debootstrap to fail in some configurations:

# debootstrap  --variant=minbase --include=cron --exclude=apt bookworm /trial
http://deb.debian.org/debian

regards,

Ralph.

Bug#1003856: fatal rendering error

[Ralph Ronnquist]

Package: debtree
Version: 1.1.0

Some graph renderings fail with the error message:
---
Can't use string ("") as a HASH ref while "strict refs" in use at
/usr/bin/debtree line 329.
---

Example: debtree g++

regards,

Ralph.

Bug#650234: More information

[Ralph Ronnquist]

Package: fakechroot
Version: 2.19-3.2

Commenting out the lines mentioning fakechroot_chroot_paths_ldsoconf
in /usr/sbin/chroot.fakechroot, and the resulting spurious "fi",
avoids this problem.

Ralph.

Bug#933697: libuuid1 declared to replace e2fsprogs

[Ralph Ronnquist]

Thanks. Though, I'm afraid I jumped to a wrong conclusion from
'apt-cache depends libuuid1' claiming to replace e2fsprogs without
showing the version caveat. It's now clearer that my problem comes up in
Devuan beowulf but not in Debian buster, and as you already pointed out,
this particular thing is rather a non-issue.

Ralph.

Theodore Y. Ts'o wrote on 3/8/19 1:46 am:
> On Fri, Aug 02, 2019 at 02:09:03PM +1000, Ralph Ronnquist wrote:
>> Well,
>>
>> when I then "agt-get install libuuid1:i386" (on this multiarch)
>> I get advice about a page full of packages to be removed, and the
>> following (plus a bit more):
>> ---
>> WARNING: The following essential packages will be removed.
>> This should NOT be done unless you know exactly what you are doing!
>>   e2fsprogs libblkid1 (due to e2fsprogs) libuuid1 (due to e2fsprogs) fdisk
>>   libfdisk1 (due to fdisk) libmount1 (due to fdisk) init
>>   sysvinit-core (due to init) mount util-linux (due to mount) sysvinit-utils
>> 
> 
> What version of e2fsprogs did you have on the system at that time?
> 
>  - Ted
> 

Bug#933697: libuuid1 declared to replace e2fsprogs

[Ralph Ronnquist]

On a second try, with a fresh live ISO, I had to mess around quite a bit
before arriving at that warning. So perhaps my previous system was a bit
too far from pristine.

Ralph.

Ralph Ronnquist wrote on 2/8/19 2:09 pm:
> Well,
> 
> when I then "agt-get install libuuid1:i386" (on this multiarch)
> I get advice about a page full of packages to be removed, and the
> following (plus a bit more):
> ---
> WARNING: The following essential packages will be removed.
> This should NOT be done unless you know exactly what you are doing!
>   e2fsprogs libblkid1 (due to e2fsprogs) libuuid1 (due to e2fsprogs) fdisk
>   libfdisk1 (due to fdisk) libmount1 (due to fdisk) init
>   sysvinit-core (due to init) mount util-linux (due to mount) sysvinit-utils
> 
> 
> Then, as I go ahead and install anyhow, I end up with a rather broken
> installation :(
> 
> So "harmless" is probably not the right word.
> 
> Ralph.
> 
> Theodore Y. Ts'o wrote on 2/8/19 1:41 pm:
>> On Fri, Aug 02, 2019 at 11:25:10AM +1000, Ralph Ronnquist wrote:
>>> Package: libuuid1
>>> Version: 2.34-0.1
>>>
>>> The package is declared to replace e2fsprogs, which it doesn't do.
>>> Rather, installing it has a fair few ramifications on the installed system.
>>>
>>> The package belongs to the util-linux source, and it seems to be the
>>> same issue with uuid-runtime and uuid-dev.
>>
>> The replaces line in question is:
>>
>> Replaces: e2fsprogs (<< 1.34-1)
>>
>> This was because libuuid1 used to shipped as part of e2fsprogs, and it
>> was split out in 1.34-1:
>>
>> e2fsprogs (1.34-1) unstable; urgency=low
>>
>>   * Split shared libraries out of the e2fsprogs package into separate
>> packages: libss2, libcomerr2, libuuid1, and e2fslibs.  (Closes: #201155,
>> #201164)
>>
>> This happened in 2003, so this split landed in Debian 3.1 (Sarge).
>> Later on, the uuid and blkid libraries got moved to util-linux, and
>> the Replaces line carried over.
>>
>> Given that Debian 10 (Buster) is now stable, that Replaces line is
>> O-B-S-O-L-E-T-E.   That being said, it's harmless.
>>
>> Cheers,
>>
>>  - Ted
>>

Bug#933697: libuuid1 declared to replace e2fsprogs

[Ralph Ronnquist]

Well,

when I then "agt-get install libuuid1:i386" (on this multiarch)
I get advice about a page full of packages to be removed, and the
following (plus a bit more):
---
WARNING: The following essential packages will be removed.
This should NOT be done unless you know exactly what you are doing!
  e2fsprogs libblkid1 (due to e2fsprogs) libuuid1 (due to e2fsprogs) fdisk
  libfdisk1 (due to fdisk) libmount1 (due to fdisk) init
  sysvinit-core (due to init) mount util-linux (due to mount) sysvinit-utils


Then, as I go ahead and install anyhow, I end up with a rather broken
installation :(

So "harmless" is probably not the right word.

Ralph.

Theodore Y. Ts'o wrote on 2/8/19 1:41 pm:
> On Fri, Aug 02, 2019 at 11:25:10AM +1000, Ralph Ronnquist wrote:
>> Package: libuuid1
>> Version: 2.34-0.1
>>
>> The package is declared to replace e2fsprogs, which it doesn't do.
>> Rather, installing it has a fair few ramifications on the installed system.
>>
>> The package belongs to the util-linux source, and it seems to be the
>> same issue with uuid-runtime and uuid-dev.
> 
> The replaces line in question is:
> 
> Replaces: e2fsprogs (<< 1.34-1)
> 
> This was because libuuid1 used to shipped as part of e2fsprogs, and it
> was split out in 1.34-1:
> 
> e2fsprogs (1.34-1) unstable; urgency=low
> 
>   * Split shared libraries out of the e2fsprogs package into separate
> packages: libss2, libcomerr2, libuuid1, and e2fslibs.  (Closes: #201155,
> #201164)
> 
> This happened in 2003, so this split landed in Debian 3.1 (Sarge).
> Later on, the uuid and blkid libraries got moved to util-linux, and
> the Replaces line carried over.
> 
> Given that Debian 10 (Buster) is now stable, that Replaces line is
> O-B-S-O-L-E-T-E.   That being said, it's harmless.
> 
> Cheers,
> 
>   - Ted
> 

Bug#933697: libuuid1 declared to replace e2fsprogs

[Ralph Ronnquist]

Package: libuuid1
Version: 2.34-0.1

The package is declared to replace e2fsprogs, which it doesn't do.
Rather, installing it has a fair few ramifications on the installed system.

The package belongs to the util-linux source, and it seems to be the
same issue with uuid-runtime and uuid-dev.

Ralph.

Bug#831382: closed by Sebastian Ramacher <sramac...@debian.org> (Re: Bug#831382: mplayer2: keyboard shortcuts fail for up/down/left/right arrows.. program hangs in "seek" (apparently))

[Ralph Ronnquist]

I'll take this as an impolite way of aking "Does this happen on Debian 8 
Jessie as well?", and the answer is "Yes, it does indeed."


On 15/07/16 20:42, Debian Bug Tracking System wrote:

This is an automatic notification regarding your Bug report
which was filed against the mplayer2 package:

#831382: mplayer2: keyboard shortcuts fail for up/down/left/right arrows.. program hangs 
in "seek" (apparently)

It has been closed by Sebastian Ramacher .

Their explanation is attached below along with your original report.
If this explanation is unsatisfactory and you have not received a
better one in a separate message then please contact Sebastian Ramacher 
 by
replying to this email.

Bug#831382: mplayer2: keyboard shortcuts fail for up/down/left/right arrows.. program hangs in "seek" (apparently)

[Ralph Ronnquist]

Package: mplayer2
Version: Jessie stable
Severity: important

Dear Maintainer,

*** Reporter, please consider answering these questions, where appropriate ***

   * What led up to the situation?
   * What exactly did you do (or not do) that was effective (or
 ineffective)?
   * What was the outcome of this action?
   * What outcome did you expect instead?

*** End of the template - remove these template lines ***


-- System Information:
Distributor ID: Devuan
Description:Devuan GNU/Linux 1.0 (jessie)
Release:1.0
Codename:   jessie
Architecture: x86_64

Kernel: Linux 3.16.0-4-amd64 (SMP w/8 CPU cores)
Locale: LANG=en_AU.UTF-8, LC_CTYPE=en_AU.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
Init: sysvinit (via /sbin/init)

I tried mplayer2 as well as mpv, and they both exhibit the same
problem, that arrow key short cuts makes the program hang. I've now
installed mplayer from deb-multimedia, and this works fine.

Bug#697854: general: Fail to display video

[Ralph Ronnquist]

I'm afraid that machine now lives its busy life as ubuntu 12.04 so I 
can't reproduce it very easily.


Though, if I can get a couple of hours access to the h/w again, I can 
probably give it a try. But, since I do have limited access to the 
machine, it would be useful if perhaps you could email me beforehand any 
test scenarios and information requests that spring to mind, for 
tracking particulars.


As for version, it was a completely fresh 6.0 install off the net at the 
time, but I didn't log the versions of packages. I didn't keep 
installation media, so I will need to try the present debian 6, and 
maybe 7 as well (especially if it still shows up on 6).


regards,

Ralph.

On 21/02/14 01:54, althaser wrote:

Hey,

Could you please still reproduce this bug ? what is the cheese version ?

thanks
regards
althaser




--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#698099: lyx: The shortcut sequence Alt-E, M, F, D for an inlined formula box doesn't work

[Ralph Ronnquist]

Package: lyx
Version: 2.0.3-3
Severity: minor

Dear Maintainer,

for some reason the shortcut sequence Alt-E,M,F,D for changing an inlined
formula box into a displayed formula box doesn't work, but the last two
keys must be repeated; i.e., Alt-E,M,F,D,F,D works, and so does the direct
shortcut Alt-M,T,D

Obviously very minor, but I had some spare time...

-- System Information:
Debian Release: 6.0.5
  APT prefers stable
  APT policy: (1000, 'stable')
Architecture: i386 (i686)

Kernel: Linux 3.2.0-2-686-pae (SMP w/8 CPU cores)
Locale: LANG=en_AU.UTF-8, LC_CTYPE=en_AU.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

Versions of packages lyx depends on:
ii  libboost-regex1.49.01.49.0-3.1
ii  libboost-signals1.49.0  1.49.0-3.1
ii  libc6   2.13-33
ii  libenchant1c2a  1.6.0-1
ii  libgcc1 1:4.7.1-2
ii  libmythes-1.2-0 2:1.2.1-1
ii  libqtcore4  4:4.8.2-1
ii  libqtgui4   4:4.8.2-1
ii  libstdc++6  4.7.1-2
ii  lyx-common  2.0.3-3
ii  xdg-utils   1.0.2+cvs20100307-2
ii  zlib1g  1:1.2.3.4.dfsg-3

Versions of packages lyx recommends:
ii  dvipng   1.14-1+b1
ii  elyxer   1.2.3-1
ii  evince [pdf-viewer]  3.4.0-2+b1
ii  fonts-lyx2.0.3-3
ii  ghostscript  9.05~dfsg-6
ii  imagemagick  8:6.7.7.2-1
ii  poppler-utils0.12.4-1.2
ii  preview-latex-style  11.86-10
ii  psutils  1.17-27
ii  texlive-fonts-recommended2012.20120611-3
ii  texlive-generic-extra2012.20120611-1
ii  texlive-generic-recommended  2012.20120611-3
ii  texlive-latex-extra  2012.20120611-1
ii  texlive-latex-recommended2012.20120611-3
ii  texlive-science  2012.20120611-1

Versions of packages lyx suggests:
pn  chktex   none
pn  dvipost  none
pn  etoolbox none
pn  gnuhtml2latexnone
pn  groffnone
pn  latex2rtfnone
pn  librsvg2-bin | inkscape  none
pn  libtiff-toolsnone
pn  linuxdoc-tools   none
pn  lyx-dbg  none
pn  nowebnone
pn  rcs  none
pn  sgmltools-lite   none
pn  texlive-xetexnone
pn  writer2latex none
pn  wv   none

-- no debconf information


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#697854: general: Fail to display video

[Ralph Ronnquist]

Package: general
Severity: important

I just installed a fresh Debian 6.0, and got the problem that video doesn't
show, specifically in skype (neither preview nor call) and cheese (preview).

The webcam appears to work fine, since it shows up at the remote end on skype
call (but I see nothing here), and cheese captures pictures nicely.

Also, when I ssh from another host with ForwardX11, then skype (both preview
and call) and cheese preview show video on that host.

Thus, I'm guessing on something with X but don't know where or what.

The host is a Sony Vaio Z with:
00:02.0 VGA compatible controller: Intel Corporation Sandy Bridge Integrated
Graphics Controller (rev 09)



-- System Information:
Debian Release: 6.0.6
  APT prefers stable
  APT policy: (900, 'stable'), (400, 'testing')
Architecture: amd64 (x86_64)

Kernel: Linux 3.2.0-0.bpo.4-amd64 (SMP w/4 CPU cores)
Locale: LANG=en_AU.UTF-8, LC_CTYPE=en_AU.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#697854: closed by Holger Levsen hol...@layer-acht.org (Re: Bug#697854: general: Fail to display video)

[Ralph Ronnquist]

Well, as far as I can see it's a problem with the X server software 
rather than anything else, or the particular collection of packages I 
got on that installation. It happens with both cheese and skype, which 
are the things I've tried, while guvcview fails to run completely, with 
the note Fatal:g_thread NOT supported' (a different problem).


Perhaps you would you be able to direct me to a better place to lodge 
this, if this is the wrong place?


regards,

Ralph.

On 11/01/13 01:09, Debian Bug Tracking System wrote:

This is an automatic notification regarding your Bug report
which was filed against the general package:

#697854: general: Fail to display video

It has been closed by Holger Levsenhol...@layer-acht.org.

Their explanation is attached below along with your original report.
If this explanation is unsatisfactory and you have not received a
better one in a separate message then please contact Holger 
Levsenhol...@layer-acht.org  by
replying to this email.





--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#697854: general: Fail to display video

[Ralph Ronnquist]

I'm afraid I have tried this; I have libv4l-0 installed, and tried the 
LD_PRELOAD variants as well. No luck :-(


The point is that it does work fine when displaying on another host via 
ssh with X11 forwarding (though skype needs LD_PRELOAD). (Possibly plain 
remote X11 as well, but I don't have that set up) This leads me to 
believe there's something with the X server subsystem.


In any case I need help to dig into it.

Ralph.

On 11/01/13 01:43, Abou Al Montacir wrote:

On Fri, 2013-01-11 at 00:44 +1100, Ralph Ronnquist wrote:

Package: general
Severity: important

I just installed a fresh Debian 6.0, and got the problem that video doesn't
show, specifically in skype (neither preview nor call) and cheese (preview).

The webcam appears to work fine, since it shows up at the remote end on skype
call (but I see nothing here), and cheese captures pictures nicely.

Also, when I ssh from another host with ForwardX11, then skype (both preview
and call) and cheese preview show video on that host.

Thus, I'm guessing on something with X but don't know where or what.

The host is a Sony Vaio Z with:
00:02.0 VGA compatible controller: Intel Corporation Sandy Bridge Integrated
Graphics Controller (rev 09)


Please install libv4l-0 and copy this launcher
# aptitude install libv4l-0
# cp skype.desktop /usr/share/applications/skype.desktop

This should fix your issue and probably all those googling for
skype+debian+webcam

Cheers,



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#697854: closed by Holger Levsen hol...@layer-acht.org (Re: Bug#697854: general: Fail to display video)

[Ralph Ronnquist]

The problem in cheese is that the preview display is black. It does snap 
from the webcam but the preview screen remains black. But, if I run 
cheese with X11 redirection to another server, then the preview display 
on it is not black. That's how it relates to cheese.


I don't know anything about the details, but I would agree that this 
probably comes up due to some kind of library mismatch in X or how 
cheese (and other programs) use X. Tonight I will test Anne Bezemer's 
solution (that I got via direct email), which is to use 
DISPLAY=localhost:0.0 rather than the default, and hopefully get it 
working in that way.


Whichever: this clearly is a bug in Debian 6.0 (on that machine). Last 
week I had Linux Mint 12 on the same machine, and not this problem;

it operated the display without workarounds.

It probably is not local to cheese, but cheese is a good and easy way of 
triggering it. Hopefully this bug can remain open, though filed in the 
right bucket, at least until I have my workaround. And then someone with 
deeper insight might make something useful of it.


regards,

Ralph.

On 11/01/13 11:50, Michael Biebl wrote:

On 10.01.2013 19:15, Holger Levsen wrote:

reopen 697854
reassign 697854 cheese
thanks
# dear cheese maintainers,
# below is some context, please read the full bug log for full context ;)

Hi Ralph,

On Donnerstag, 10. Januar 2013, Ralph Ronnquist wrote:

Well, as far as I can see it's a problem with the X server software
rather than anything else, or the particular collection of packages I
got on that installation. It happens with both cheese and skype, which
are the things I've tried, while guvcview fails to run completely, with
the note Fatal:g_thread NOT supported' (a different problem).

Perhaps you would you be able to direct me to a better place to lodge
this, if this is the wrong place?


as this bug also occurs with cheese, I've reopened and reassigned it there.
Thanks for your help in tracking this error down!



Why should a problem in skype be a symptom for a cheese problem? We
rather re-assign that to general or close it again. The bug report is
not really useful as is.





--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#697854: closed by Holger Levsen hol...@layer-acht.org (Re: Bug#697854: general: Fail to display video)

[Ralph Ronnquist]

Ok; No luck with X11 redirection either, and now it segfaults. The 
following is from my run attempt of cheese under gdb:

-
Reading symbols from /usr/bin/cheese...(no debugging symbols found)...done.
(gdb) r
Starting program: /usr/bin/cheese
[Thread debugging using libthread_db enabled]
[New Thread 0x7fffec017700 (LWP 3483)]
[Thread 0x7fffec017700 (LWP 3483) exited]
[New Thread 0x7fffec017700 (LWP 3484)]
[Thread 0x7fffec017700 (LWP 3484) exited]
[New Thread 0x7fffec017700 (LWP 3485)]
[Thread 0x7fffec017700 (LWP 3485) exited]
[New Thread 0x7fffec017700 (LWP 3486)]
[New Thread 0x7fffe3fff700 (LWP 3487)]
[New Thread 0x7fffe37fe700 (LWP 3488)]
[New Thread 0x7fffe2ffd700 (LWP 3489)]
[New Thread 0x7fffe27fc700 (LWP 3490)]
[Thread 0x7fffec017700 (LWP 3486) exited]
[New Thread 0x7fffec017700 (LWP 3494)]
[Thread 0x7fffec017700 (LWP 3494) exited]
The program 'cheese' received an X Window System error.
This probably reflects a bug in the program.
The error was 'BadAlloc (insufficient resources for operation)'.
  (Details: serial 77 error_code 11 request_code 132 minor_code 19)
  (Note to programmers: normally, X errors are reported asynchronously;
   that is, you will receive the error a while after causing it.
   To debug your program, run it with the --sync command line
   option to change this behavior. You can then get a meaningful
   backtrace from your debugger if you break on the gdk_x_error() 
function.)

[Thread 0x7fffe2ffd700 (LWP 3489) exited]
[Thread 0x7fffe27fc700 (LWP 3490) exited]
[Thread 0x7fffe37fe700 (LWP 3488) exited]
[Thread 0x7fffe3fff700 (LWP 3487) exited]

Program exited with code 01.


This machine is hereby being demoted to Ubuntu 12.04 32bit (though with 
xfce rather than gnome)  fingers crossed :-)


cheers,

Ralph.

On 11/01/13 11:50, Michael Biebl wrote:

On 10.01.2013 19:15, Holger Levsen wrote:

reopen 697854
reassign 697854 cheese
thanks
# dear cheese maintainers,
# below is some context, please read the full bug log for full context ;)

Hi Ralph,

On Donnerstag, 10. Januar 2013, Ralph Ronnquist wrote:

Well, as far as I can see it's a problem with the X server software
rather than anything else, or the particular collection of packages I
got on that installation. It happens with both cheese and skype, which
are the things I've tried, while guvcview fails to run completely, with
the note Fatal:g_thread NOT supported' (a different problem).

Perhaps you would you be able to direct me to a better place to lodge
this, if this is the wrong place?


as this bug also occurs with cheese, I've reopened and reassigned it there.
Thanks for your help in tracking this error down!



Why should a problem in skype be a symptom for a cheese problem? We
rather re-assign that to general or close it again. The bug report is
not really useful as is.





--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#684111: wicd gui doesn't update pre/post connection script settings

[Ralph Ronnquist]

Package: wicd
Version: 1.7.2.4-2
Severity: normal
Tags: patch

Dear Maintainer,

I had the problem that the wicd GUI did not keep any changes that I
made to the pre/post connection script settings, via the property
window for an interface.

After some digging, it appears that the ConsoleManager interface
got changed to have write() without arguments, so I patched my
/usr/share/wicd/gtk/configscript.py for this, according to the
following diff:
-
95c95,96
 con.write(open(wired_conf, w))
---
 #con.write(open(wired_conf, w))
 con.write()
106c107,108
 con.write(open(wireless_conf, w))
---
 #con.write(open(wireless_conf, w))
 con.write()
-

That seems to have made it work for me.

regards,

Ralph.

-- System Information:
Debian Release: 6.0.5
  APT prefers testing
  APT policy: (1001, 'testing'), (1000, 'stable'), (500, 'stable-updates')
Architecture: i386 (i686)

Kernel: Linux 3.2.0-2-686-pae (SMP w/8 CPU cores)
Locale: LANG=en_AU.UTF-8, LC_CTYPE=en_AU.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

Versions of packages wicd depends on:
ii  wicd-daemon 1.7.2.4-2
ii  wicd-gtk [wicd-client]  1.7.2.4-2

wicd recommends no packages.

wicd suggests no packages.

Versions of packages wicd-gtk depends on:
ii  python 2.7.3~rc2-1
ii  python-glade2  2.24.0-3
ii  python-gtk22.24.0-3
ii  wicd-daemon1.7.2.4-2

Versions of packages wicd-gtk recommends:
ii  gksu   2.0.2-5
ii  python-notify  0.1.1-3

Versions of packages wicd-daemon depends on:
ii  adduser 3.112+nmu2
ii  dbus1.2.24-4+squeeze1
ii  debconf 1.5.36.1
ii  iproute 20100519-3
ii  iputils-ping3:20100418-3
ii  isc-dhcp-client [dhcp3-client]  4.1.1-P1-15+squeeze3
ii  lsb-base3.2-23.2squeeze1
ii  net-tools   1.60-23
ii  psmisc  22.11-1
ii  python  2.7.3~rc2-1
ii  python-dbus 1.1.0-1
ii  python-gobject  3.2.2-1
ii  python-wicd 1.7.2.4-2
ii  wireless-tools  30~pre9-8
ii  wpasupplicant   0.6.10-2.1

Versions of packages wicd-daemon recommends:
ii  wicd-gtk [wicd-client]  1.7.2.4-2

Versions of packages wicd-daemon suggests:
ii  pm-utils  1.3.0-3

Versions of packages python-wicd depends on:
ii  python 2.7.3~rc2-1
ii  python2.6  2.6.6-8+b1
ii  python2.7  2.7.3~rc2-2.1

-- debconf information:
* wicd/users:


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#681646: Test example

[Ralph Ronnquist]

The following is an example of this bug by a very small session, after 
having removed ~/.gnuplot_history before starting gnuplot. I have added 
my own numbering of the commands.


1: gnuplot a=5;plot x
2: gnuplot plot x
3: gnuplot history
1  history
4: gnuplot plot x
5: gnuplot plot x
6: gnuplot plot 2*x
7: gnuplot history
1  history
2  plot 2*x
3  history
8: gnuplot plot x
9: gnuplot q


Command #1 was typed as is.

Command #2 was obtained by up-arrow, 6*left-arrow, 4*rubout, enter.

Command #3 was typed as is; before typing, up-arrow recall showed me 
command #1 and not #2. But neither are in the history (file).


Command #4 was typed as is; recall before typing showed #3 and #1, not #2.

Command #5 was typed as is; recall before typing shows #3 and #1, not #2 
or #4.


Command #6 was typed as is; recall before typing shows #3 and #1, not 
#2, #4 or #5.


Command #7 was typed as is; recall before typing shows #6, #3 and #1, 
not #2, #4 or #5.


Command #8 was typed as is; recall before typing shows #7, #6, #3 and 
#1, not #2, #4 or #5.


Command #9 was typed as is; recall before typing shows #7, #6, #3 and 
#1, not #2, #4, #5 or #8.



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#681646: gnuplot: somtimes forgets previous command (can't see it in history either)

[Ralph Ronnquist]

Package: gnuplot
Version: 4.6.0-8
Severity: normal

Dear Maintainer,

now and then gnuplot seems to fail to store the command in its history,
so I can't recall it by ^P or up-arrow or any history command. In my
typical use, I recall previous command, make an edit and push enter.
It usually works fine, but sometimes gnuplots fails to remember the
command, so on next recall, I instead get a previous (what I did 2
command lines ago).

When it start, gnuplot insists on not remembering that same command line,
while it happily remembers other command lines in between, whether I
retype it manually, or recall it from history way back.

The following is a snapshot of gnuplot history:

  635  wx=0;w=12;t=t2;h=800;load plot1
  636  wx=0;w=12;t=t2;h=800;load plot1
  637  wx=0;w=12;t=t2;h=800;load plot1
  638  wx=0;w=8;t=t2;h=800;load plot1
  639  wx=0;w=8;t=6000;h=800;load plot1
  640  wx=0;w=8;t=6000;h=800;load plot1

Here I actually tried a callback form #639, to reuse #638, but that didn't
work, so I retyped #638, which is not seen in the history. Then I used 
recall (up-arrow) from a earlier command (way up), which it remembered
as #639. Then I tried again by retyping #638, which also did not go into
the history, and #640 is a subsequent recall of #639.

When it happens, that command line cannot be entered into the history,
if I add two spaces to the bigging it gets remembered.

It happens often enough to be a bother, but not often enough to be
a nuisance :-) and I haven't worked out a more detailed case pattern.
Possibly it has something to do with that the line first is by a
recall where I deleted the first few characters (up to a semicolon).

-- System Information:
Debian Release: 6.0.5
  APT prefers testing
  APT policy: (1001, 'testing'), (1000, 'stable'), (500, 'stable-updates')
Architecture: i386 (i686)

Kernel: Linux 3.2.0-2-686-pae (SMP w/8 CPU cores)
Locale: LANG=en_AU.UTF-8, LC_CTYPE=en_AU.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

Versions of packages gnuplot depends on:
ii  gnuplot-x11 [gnuplot-nox]  4.6.0-8

gnuplot recommends no packages.

Versions of packages gnuplot suggests:
pn  gnuplot-doc  none

-- no debconf information


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org