: Ramkumar Ramachandra
---
units/sys-kernel-debug.mount | 1 +
1 file changed, 1 insertion(+)
diff --git a/units/sys-kernel-debug.mount b/units/sys-kernel-debug.mount
index 5369728..f90bfc1 100644
--- a/units/sys-kernel-debug.mount
+++ b/units/sys-kernel-debug.mount
@@ -17,3 +17,4 @@ Before
Ramkumar Ramachandra wrote:
> Got it. So, it should have provided /dev/tum* (or something) instead
> of /dev/tty*. But what can we do about it now? I certainly cannot
> remove /dev/tty* and murder the users.
Wait, I think I have an idea to preserve backward compatibility. I'll
r
Lennart Poettering wrote:
> So please, ask the UML guys to somehow cleanup their VT mess. It's their
> broken code, and their job to fix it.
The question is how? Just tell me what to do, and I'll do it myself.
> I can understand if you are disappointed by this, but please try to see
> it from ou
Lennart Poettering wrote:
> They can call their ttys any way they want. If the call them
> /dev/tty[1..64] however, then they need to implement the VC
> interfaces. All of them.
>
> Also note that some hypervisors implement /dev/hvc0, /dev/xvc0,
> /dev/hvsi0 and suchlike. Theser are also ttys, whic
Kay Sievers wrote:
> [...]
Fwiw, I more-or-less agree with the principle "don't accumulate
specific workarounds when it's not scalable or maintainable".
> Hacks like the UML tty0 hack is just a too broken idea, to
> get any support or an exception in a generic tool like systemd.
I think of it as
Kay Sievers wrote:
> You meant to provide a patch against git, not your earlier patches?
Ofcourse; I was just asking if this is what you meant before
submitting another iteration for inclusion.
___
systemd-devel mailing list
[email protected]
Kay Sievers wrote:
> Where do you see any "specific exceptions for broken hacks"?
>
> We surely support different forms of virtualizations, and support
> reasonable custom behavior. But we do not support providing a tty0 and
> have no working VT. It's just madness we don't want to see anywhere,
> a
Kay Sievers wrote:
> No, sorry, the same rule applies to C code as it does to unit files:
> We do not collect specific exceptions for broken hacks in exotic
> tools. Means: there will be _zero_ matches on UML strings in systemd
Um, then why do we have a detect-virt at all? A quick grep tells me
t
Hi,
I noticed that my TERM is set to "linux", even when I boot the
console-getty.service. Isn't it supposed to be "console" instead,
since VT emulation is not guaranteed? Should we explicitly pass the
term argument to agetty?
Thanks.
___
systemd-devel
indicate
that /dev/tty0 is not a virtual console.
Cc: Kay Sievers
Cc: Zbigniew Jędrzejewski-Szmek
Cc: Lennart Poettering
Signed-off-by: Ramkumar Ramachandra
---
I didn't cut corners: I built it with abs, and tested it properly :)
Kay: You like this?
src/vconsole/vconsole-setup.c | 5 +++
Kay Sievers wrote:
>> If you're so particular about keeping primary unit files "clean", may
>> I suggest moving the exception code to vconsole-setup.c?
>
> That's what I meant, yeah, sorry for the confusion.
Cool, I understand that your desire to keep primary units clean: I'll
write a patch for th
Kay Sievers wrote:
> such work-arounds need to
> be a generic as possible
I disagree with this. I specifically asked Lennart about a /dev/tty0
being created by hand using mknod earlier: his answer was that systemd
does not support it. In the most generic scenario, you cannot rely on
the sanity o
Kay Sievers wrote:
> Oh, I meant turning is_vconsole() into returning success when it is
> called for a tty0 that in fact doesn't work like tty0. :)
We might as well remove is_vconsole() then. Does it make sense to
execute the rest of the code in vconsole-setup.c (fontmap, utf8 etc.)
on a non-VT
Kay Sievers wrote:
> It's not about failing it, that would be ugly, sure. Vconsole can just
> give up and return 0 if it finds a tt0 that actually isn't a tty0.
Hm, how can I tell if vconsole-setup actually did something or not
then? Doesn't it matter for curses apps, which require VT102
emulatio
Kay Sievers wrote:
> I would rather see vconsole to detect the mess and silently give up,
> instead of adding exotic options for really weird faked and wrong tty0
> setups.
um Linux lacks VT102 emulation, and this is a documented fact.
Therefore, your code:
unsigned char data[1];
lay a prompt immediately on the same terminal, along with the
boot messages. This is much more user-friendly, than making the user
explicitly connect to the right device using
$ screen /dev/pts/8
to get a login prompt.
Cc: Thomas Bächler
Cc: Lennart Poettering
Signed-off-by: Ramkumar Ramach
Hi,
These are the finishing touches required to make User-Mode Linux work
out-of-the-box. Sorry it took so long: I had to build with abs and
test it for good.
Thanks again, for all the help.
Ramkumar Ramachandra (2):
getty-generator: run console shell for uml
vconsole-setup: setup negative
User-Mode Linux does not do VT102 emulation, and
$ vconsole-setup /dev/tty0
will always fail on it. In order to prevent vconsole-setup.service from
always failing, disable it when uml is detected.
Cc: Zbigniew Jędrzejewski-Szmek
Cc: Lennart Poettering
Signed-off-by: Ramkumar Ramachandra
Lennart Poettering wrote:
> /dev/tty0 always points to the foreground VC tty. Are you saying UML
> provides /dev/tty0 but not /dev/tty1, /dev/tty2, ...? I am pretty sure
> that UML should be fixed to provide neither or both, evertyhing else is
> broken.
Thanks for all the information. With some h
Lennart Poettering wrote:
> /dev/tty0 only exists
> on kernels with a VT subsystem
Yeah, but um Linux boots fine when I do
$ mv
/etc/systemd/system/getty.target.wants/{[email protected],[email protected]}
I'm reading arch/um/drivers/tty.c, but I'm not sure I understand what
exactly is going
Lennart Poettering wrote:
> Similar story here. systemd-vconsole-setup is actually conditionalized
> on ConditionPathExists=/dev/tty0, i.e. will only run on systems where
> the VT is actually built into the kernel, and skipped on all others. If
> UML has not useful VT I'd recommend compiling it out
system-vconsole-setup.service runs systemd-vconsole-setup to set up
virtual consoles, but console-getty.service implies that virtual
consoles aren't present, and that we're falling back to /dev/console.
Cc: Lennart Poettering
Signed-off-by: Ramkumar Ramachandra
---
Noticed on um linu
Ramkumar Ramachandra wrote:
> It seems that UM does not have a proper tty system; I force
> systemd to use the /dev/console by doing:
>
> $ mv
> /etc/systemd/system/getty.target.wants/{[email protected],[email protected]}
Where should the exception for um linux go? When d
-9-ge5fd680 (...)
bogomips: 7007.43
So, grep for the string "\nvendor_id\t: User Mode Linux\n" in
/proc/cpuinfo, and say "uml" when asked.
Cc: Lennart Poettering
Signed-off-by: Ramkumar Ramachandra
---
Minor tweaks suggested by Lennart. Tested.
man/systemd-
-9-ge5fd680 (...)
bogomips: 7007.43
So, grep for the string "vendor_id\t: User Mode Linux\n" in
/proc/cpuinfo, and say "uml" when asked.
Cc: Lennart Poettering
Signed-off-by: Ramkumar Ramachandra
---
See arch/um/kernel/um_arch.c:75, where it writes the hard-coded
-9-ge5fd680 (...)
bogomips: 7007.43
So, grep for the string "User Mode Linux" in /proc/cpuinfo, and say
"uml" when asked.
Cc: Lennart Poettering
Signed-off-by: Ramkumar Ramachandra
---
It seems that UM does not have a proper tty system; I force
systemd to use
David Strauss wrote:
> We could also move more tools outside the main systemd package and
> into PIP.
Er, you want to move some essential systemd scripts out from the source tree?
Ram
___
systemd-devel mailing list
[email protected]
ht
Hi Michael,
Michael Biebl wrote:
> 2012/12/2 Ramkumar Ramachandra :
>> Hm, you have a point. On Debian, "/etc/machine-id" is created when
>> the systemd package is installed -- I'm not sure if it's in the source
>> code or if it's a custom Debian
Zbigniew Jędrzejewski-Szmek wrote:
> I started munging your patch to apply it, but on second thought it is
> totally the wrong direction to take. The purpose of tests is to check
> if systemd will function after installation. And as mentioned
> elsewhere in the thread, journald will break badly, an
David Strauss wrote:
> Would it be possible, for testing purposes, to generate a machine ID
> on the fly if one is not present on disk?
Hard, as "/etc/machine-id" is hardcoded in
src/libsystemd-id128/sd-id128.c. I don't think it's worth the effort.
Ram
___
The following tests fail if /etc/machine-id is not present:
$ ./test-id128
random: a08ea8ed34594d4bbd953dd182ec86f9
Assertion 'sd_id128_get_machine(&id) == 0' failed at
src/test/test-id128.c:41, function main(). Aborting.
[1]8017 abort (core dumped) ./test-id128
$ ./test-journal
Ramkumar Ramachandra wrote:
> I booted up my Debian machine with systemd, and removed systemd after
> bootup. Now it refuses to halt with:
>
> init: timeout opening/writing control channel /dev/initctl
Correction: I can't halt a Debian systemd machine at all (even when
Hi,
I booted up my Debian machine with systemd, and removed systemd after
bootup. Now it refuses to halt with:
init: timeout opening/writing control channel /dev/initctl
I don't know if it's systemd's fault or the Debian package manager's
fault- investigating.
Ram
___
David Strauss wrote:
> Please fix the spelling of "present."
> The number -128, if used, should get #define'd instead of being a magic
> number dropped in a bunch of places.
Thanks. Will fix in the next iteration.
> It's not good to mask all errors opening the machine-id with the -128
> response
Michael Biebl wrote:
> 2012/11/30 Ramkumar Ramachandra :
>> Execute python using "/usr/bin/env python" instead of hard-coding the
>
> I'm not really a fan of using /usr/bin/env, as you can pick up a
> random python version depending on how your PATH is setup
The following tests fail if /etc/machine-id is not present:
$ ./test-id128
random: a08ea8ed34594d4bbd953dd182ec86f9
Assertion 'sd_id128_get_machine(&id) == 0' failed at
src/test/test-id128.c:41, function main(). Aborting.
[1]8017 abort (core dumped) ./test-id128
$ ./test-journal
is is because they call sd_id128_get_machine() which barfs if
/etc/machine-id can't be open()'ed. Treat this as a special case and
skip the dependent tests instead of failing them.
---
Zbigniew Jędrzejewski-Szmek wrote:
> On Fri, Nov 30, 2012 at 06:47:42PM +0530, Ramkumar
Kay Sievers wrote:
> On Thu, Nov 29, 2012 at 8:06 PM, Ramkumar Ramachandra
> wrote:
>> Update the license with the copy found on
>> http://www.gnu.org/licenses/old-licenses/lgpl-2.1.txt
>
> Applied.
Thanks. I noticed that you dropped the signoff on all three patch
Kay Sievers wrote:
> On Fri, Nov 30, 2012 at 1:45 PM, Ramkumar Ramachandra
> wrote:
>> Where is the creation code? The systemd-machine-id-setup binary does
>> _not_ create it, and my patch proposes to fix that.
>
> We overmount an empty file if needed. We do not want to
is is because they call sd_id128_get_machine() which barfs if
/etc/machine-id can't be open()'ed. Treat this as a special case and
skip the dependent tests instead of failing them.
Signed-off-by: Ramkumar Ramachandra
---
Ramkumar Ramachandra wrote:
> Kay Sievers wrote:
>> On F
Kay Sievers wrote:
> It is mandatory for systemd systems. We just require it to be around,
> and do not work around it.
Specifically, which package installs /etc/machine-id? Is it a build
dependency for systemd?
Ram
___
systemd-devel mailing list
syste
Kay Sievers wrote:
> On Fri, Nov 30, 2012 at 1:28 PM, Zbigniew Jędrzejewski-Szmek
> wrote:
>
>> It'll be created by systemd during init if not present. So we _do_ a
>> kind of around.
Where is the creation code? The systemd-machine-id-setup binary does
_not_ create it, and my patch proposes to f
Ramkumar Ramachandra wrote:
> $ ./test-journal
> Assertion 'journal_file_open("test.journal", O_RDWR|O_CREAT, 0666,
> true, true, NULL, NULL, NULL, &f) == 0' failed at
> src/journal/test-journal.c:46, function main(). Aborting.
> [1]8059 abort (core dumpe
: Ramkumar Ramachandra
---
Ramkumar Ramachandra wrote:
> $ ./test-id128
> random: a08ea8ed34594d4bbd953dd182ec86f9
> Assertion 'sd_id128_get_machine(&id) == 0' failed at
> src/test/test-id128.c:41, function main(). Aborting.
> [1]8017 abort (core dumped)
Execute python using "/usr/bin/env python" instead of hard-coding the
path of the Python binary.
Signed-off-by: Ramkumar Ramachandra
---
Minor fix.
src/analyze/systemd-analyze |2 +-
src/locale/generate-kbd-model-map |2 +-
test/rule-syntax-check.py |2 +
Fix some minor spelling mistakes in the manpages.
Signed-off-by: Ramkumar Ramachandra
---
man/daemon.xml |2 +-
man/journalctl.xml |2 +-
man/journald.conf.xml |4 ++--
man/kernel-command-line.xml |2 +-
4 files changed, 5 insertions(+), 5 deletions
f33d3ec1 (move more common files to shared/ and add them to shared.la,
2012-04-12) moved src/hostname-setup.c to src/shared/hostname-setup.c,
but did not update DISTRO_PORTING accordingly. Fix this.
Signed-off-by: Ramkumar Ramachandra
---
DISTRO_PORTING |6 +++---
1 files changed, 3
Update the license with the copy found on
http://www.gnu.org/licenses/old-licenses/lgpl-2.1.txt
Signed-off-by: Ramkumar Ramachandra
---
LICENSE.LGPL2.1 | 81 +-
1 files changed, 38 insertions(+), 43 deletions(-)
diff --git a/LICENSE.LGPL2.1
Hi,
This evening, I was curious to know whether I could install systemd on
my Ubuntu laptop. Incidentally, I got a chance to look at the code
and fixed up a few minor things. Hope it's useful.
Ram
Ramkumar Ramachandra (3):
LICENSE.LGPL2.1: update license
DISTRO_PORTING: correct loc
Hi,
The following tests fail. Is it just me, or does this indicate work
in progress?
$ ./test-id128
random: a08ea8ed34594d4bbd953dd182ec86f9
Assertion 'sd_id128_get_machine(&id) == 0' failed at
src/test/test-id128.c:41, function main(). Aborting.
[1]8017 abort (core dumped) ./test-id128
$
Zbigniew Jędrzejewski-Szmek wrote:
> On Wed, Dec 14, 2011 at 11:19:13PM +0530, Ramkumar Ramachandra wrote:
>> Hi,
>>
>> On Dec 2, 2011, Ramkumar Ramachandra wrote:
>> > Ramkumar Ramachandra (3):
>> > Debian: update m4 scripts in units/
>> >
Hi,
On Dec 2, 2011, Ramkumar Ramachandra wrote:
> Ramkumar Ramachandra (3):
> Debian: update m4 scripts in units/
> completion: fix typo in accessing array index
> completion: add ZSH support
Does the series look alright? Can I do something to help move it forward?
Tha
Remove spurious '}'. This error went unnoticed so far because Bash
doesn't complain.
Signed-off-by: Ramkumar Ramachandra
---
src/systemd-bash-completion.sh |4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/systemd-bash-completion.sh b/s
ZSH. However, the bashcompinit feature in
ZSH implements this complete(): use it.
Signed-off-by: Ramkumar Ramachandra
---
src/systemd-bash-completion.sh | 50 ---
1 files changed, 26 insertions(+), 24 deletions(-)
diff --git a/src/systemd-bash-completion.sh
Teach the console, getty, and serial m4 scripts about Debian.
Signed-off-by: Ramkumar Ramachandra
---
units/console-shell.service.m4 |3 +++
units/[email protected]|3 +++
units/[email protected] |3 +++
3 files changed, 9 insertions(+), 0 deletions(-)
diff --git a
mail.com>
Ramkumar Ramachandra (3):
Debian: update m4 scripts in units/
completion: fix typo in accessing array index
completion: add ZSH support
src/systemd-bash-completion.sh | 54 ---
units/console-shell.service.m4 |3 ++
units/[email protected]
Hi Tollef,
Tollef Fog Heen wrote:
> | diff --git a/src/hostname-setup.c b/src/hostname-setup.c
> | index 2c2f10c..918525d 100644
> | --- a/src/hostname-setup.c
> | +++ b/src/hostname-setup.c
> | @@ -32,6 +32,8 @@
> [...]
> Not needed, the top of read_hostname from the same file.
and
> | diff --g
Teach systemd about the location of the hostname, distribution version
file, and tweak the various m4 scripts in units/.
Signed-off-by: Ramkumar Ramachandra
---
Started using systemd on my Debian machine yesterday- here's a quick
patch for starters.
Thanks for reading.
[Please CC
58 matches
Mail list logo
