On Thu, March 5, 2009 8:43 pm, aelmahmo...@users.sourceforge.net wrote:
On Sat, Feb 21, 2009 at 04:56:46PM -0800, Corey Hickey wrote:
Ok, here's the patch. Feel free to tweak it as you see fit.
I'm at my mom's house right now, so I hooked up the land line, dialed,
and
gave the patch a
On Sat, Feb 21, 2009 at 04:56:46PM -0800, Corey Hickey wrote:
Ok, here's the patch. Feel free to tweak it as you see fit.
I'm at my mom's house right now, so I hooked up the land line, dialed, and
gave the patch a rudimentary test with actual connectivity. It seems to
work fine, even with
On Wed, February 18, 2009 10:31 pm, Corey Hickey wrote:
Additionally, I can write a patch that checks the current limit and
makes slmodemd warn or die if the limit is judged to be too low. I'm
mildly in favor of that approach rather than having slmodemd modify the
limit itself.
Ok, here's the
Hello,
On Wed, Feb 18, 2009 at 01:54:38AM +0100, Mau wrote:
I've made some quick tests and, since now, the memory slmodemd allocates
never exceeded 4192KB: this seems to confirm what Corey said (4208KB,
the memory slmodemd allocates at start). Assuming what I wrote was
right, 8MB would be a
أحمد المحمودي wrote:
It sounds reasonable. As you see, I have asked Ian to comment on this.
In the meantime I see that wether we leave it unlimited or if we limit
it to say 16 MB, there is the issue that a user who may want to run
slmodemd manually (like Corey does), so we can do
Hello,
On Tue, Feb 17, 2009 at 01:51:12AM +0100, Mau wrote:
I'd try to apply your patch, which seems just fine: I think 8 MB is a
reasonable value to start with.
Ahmed, do you agree?
Maurizio
---end quoted text---
Thanks guys, I really don't have the knowledge to comment on this,
أحمد المحمودي wrote:
Yet I have been asking myself since yesterday, why does it happen with
Corey, yet not with me nor Mau ?! That problem does not seem to be
device specific problem, does it ?
I would guess that your locked memory limit is either high or unlimited.
See 'ulimit -l' or
On Tue, Feb 17, 2009 at 03:15:03AM -0800, Corey Hickey wrote:
I would guess that your locked memory limit is either high or unlimited.
See 'ulimit -l' or 'ulimit -a'.
---end quoted text---
I just found out that /etc/init.d/sl-modem-daemon got this:
ulimit -Hl unlimited; ulimit -Sl
أحمد المحمودي wrote:
On Tue, Feb 17, 2009 at 03:15:03AM -0800, Corey Hickey wrote:
I would guess that your locked memory limit is either high or unlimited.
See 'ulimit -l' or 'ulimit -a'.
---end quoted text---
I just found out that /etc/init.d/sl-modem-daemon got this:
ulimit -Hl
Just a consideration.
If I'm not wrong, assigning unlimited locked memory can give the process
the ability to lock the whole machine: dropping the privileges loses
then part of its sense.
I don't have enough knowledge to say more - I'm not even sure what I
wrote makes sense - but I feel that, if
Il 16/02/2009 00:40, Corey Hickey ha scritto:
Ok, here's the problem.
modem_main.c:976 function modem_main()
--
if (need_realtime) {
struct sched_param prm;
Maurizio Avogadro wrote:
Il 16/02/2009 00:40, Corey Hickey ha scritto:
Ok, here's the problem.
modem_main.c:976 function modem_main()
--
if (need_realtime) {
struct sched_param prm;
Il 17/02/2009 00:05, Corey Hickey ha scritto:
[...]
That's a good idea, and I tried it, but I think that's not quite how
limits.conf works. I haven't found anything that says this plainly, but
I'm pretty sure the following are true:
* Limits are applied at login by the pam_limits module,
Maurizio Avogadro wrote:
Il 13/02/2009 00:14, Corey Hickey ha scritto:
There aren't any flags on the open() system call; O_CREAT would be
needed.
The data file is actually created on the first write:
modem/modem_datafile.c:110, function datafile_save_info()
Corey Hickey wrote:
I don't know why the malloc() fails. I went to the same line in my
slmodemd binary that doesn't have 10_drop_privileges.diff, and that
malloc() succeeded, with the same arguments.
Going back to the regular privilege-dropping binary, I used gdb to run
some malloc() tests
Il 13/02/2009 00:14, Corey Hickey ha scritto:
[...]
It appears that slmodemd isn't creating /var/lib/slmodem/data.modem:1
even though it should have permissions to create files in the directory
(isn't it still running as root at that point, anyway?). Actually, now
that I check the source,
Corey Hickey wrote:
I just tried the Debian 2.6.26-1-686 kernel, and the slmodemd process
was killable, but the modem didn't work: it kept responding NO CARRIER
immediately after dialing.
[cut]
I'm going to try to figure out what it is in my particular kernel
configuration that's causing
- Forwarded message from Maurizio Avogadro mav...@gmail.com -
sorry for my late but I'm very busy... after the first glance:
1. I noticed that some previous version left a
- ---Sr--r-T root root data.(slamr|modem:)[0-9]
file in /var/lib/slmodem: maybe the slmodemd daemon is unable to
أحمد المحمودي wrote:
- Forwarded message from Maurizio Avogadro mav...@gmail.com -
sorry for my late but I'm very busy... after the first glance:
1. I noticed that some previous version left a
- ---Sr--r-T root root data.(slamr|modem:)[0-9]
file in /var/lib/slmodem: maybe the
I just tried the Debian 2.6.26-1-686 kernel, and the slmodemd process
was killable, but the modem didn't work: it kept responding NO CARRIER
immediately after dialing.
In case it's of any use, I'm attaching a couple log files:
slmodemd_2.6.26-1-686.log
slmodemd_2.6.26-1-686_no-drop-privs.log
On Sun, January 18, 2009 11:31 am, aelmahmo...@users.sourceforge.net wrote:
On Sat, Jan 17, 2009 at 01:07:48PM -0800, Corey Hickey wrote:
The process being unkillable is rather worrisome to me, and makes me
wonder if the privilege dropping is actually exposing a kernel bug. What
kernel version
On Sat, Jan 17, 2009 at 01:07:48PM -0800, Corey Hickey wrote:
The process being unkillable is rather worrisome to me, and makes me
wonder if the privilege dropping is actually exposing a kernel bug. What
kernel version are you using? I can try testing that and see if I still
have the same
Hello,
I cannot reproduce the same problem that you have encountered. I have
a question though. Does that problem happen in older versions of
sl-modem package (2.9.9d+e-pre2-10 or later) ?
If you can build binary package from source, then please try the package I
prepared at:
On Sat, January 17, 2009 2:53 am, Ø£ØÙ
د اÙÙ
ØÙ
Ùد٠wrote:
Hello,
I cannot reproduce the same problem that you have encountered. I have
a question though. Does that problem happen in older versions of
sl-modem package (2.9.9d+e-pre2-10 or later) ?
I got the source for
Package: sl-modem-daemon
Version: 2.9.11~20080817-1
Severity: important
Hello,
slmodemd from the Debian package doesn't work for me, and I can't seem to kill
the process, even with kill -9. When I compile version 2.9.11 from upstream,
it works fine. I tracked the problem down to a particular
25 matches
Mail list logo
