| --- Log opened Wed Apr 23 00:00:17 2014 | ||
| stekern | blueCmd: ah, I had already found them, but thanks for the hint | 03:39 |
|---|---|---|
| stekern | I couldn't find anything wrong with them, so have to be a problem with the kernel. I think that the kregs is not the same as the exception pt_regs | 03:41 |
| stekern | because the kregs and userregs are oif course for the cloned thread, not the one that is running | 03:42 |
| stekern | for a comparison, I'm not using any kregs at all in my eco32 port: https://github.com/skristiansson/eco32-linux/blob/master/arch/eco32/kernel/process.c#L75 | 03:47 |
| stekern | I just use a sepereate cpu_context struct to hold the context for switch_to | 03:48 |
| stekern_ | in fact, it's the userregs that hold the current saved context from exception entry | 04:07 |
| stekern_ | copied from current_pt_regs() | 04:08 |
| stekern_ | so, it should be: userregs->gpr[10] = userregs->gpr[7] | 04:09 |
| stekern | oho! that stekern_ was a handsome devil! =) | 04:09 |
| stekern | blueCmd: when I said, "I don't think that will work" I meant for the or1k port. Not sure if that might work for other archs. Of course the circumstances for it to even work without -fPIC have to be there then. | 05:06 |
| stekern | blueCmd: the ifconfig stack smashing was probably due to the struct padding too, that works now too | 05:29 |
| blueCmd | stekern: ah, I see. weird that it worked for me (or not horribly crashed at least) | 07:39 |
| blueCmd | i did rewrite that patch, maybe the first version was ok and the latter not | 07:39 |
| blueCmd | yeah, that must be it. the first version was: | 07:41 |
| blueCmd | + if (clone_flags & CLONE_SETTLS) | 07:41 |
| blueCmd | + childregs->gpr[10] = regs->gpr[7]; | 07:41 |
| stekern | ah well, it can't be that common that clone is called with CLONE_SETTLS, since everything has worked for me until I tried running irssi | 07:43 |
| stekern | ...and there's no 'regs' in copy_thread, so that first verison wouldn't compile ;) | 07:45 |
| stekern | nor childregs, was that located somewhere else then? | 07:46 |
| blueCmd | stekern: it was for an old version of linux | 07:47 |
| blueCmd | that function looked very different then | 07:47 |
| blueCmd | stekern: any more issues you have found, or is it usable as an IRC box now? | 07:48 |
| stekern | ah, yes I see, then pt_regs *regs was passed to copy_thread | 07:50 |
| stekern | and now that have been exchanged with obtaining it from current_pt_regs() instead | 07:50 |
| stekern | I wanted to irc as another user than root, but it won't let me login as anything else | 07:51 |
| stekern | 'strace su stefan' finds this: openat(AT_FDCWD, "/proc/self/loginuid", O_RDONLY) = -1 ENOENT (No such file or directory) | 07:52 |
| ysionneau | come on, root is even better, you get @ on every channels ;) | 08:22 |
| stekern | heh ;) | 08:36 |
| blueCmd | stekern: hm, I'm running as another user in qemu | 08:55 |
| blueCmd | stekern: ah, maybe loginuid is a debian patch thing | 08:56 |
| blueCmd | stekern: if you could try to track the problem down that would be cool :) | 08:57 |
| stekern | the kernel has some CONFIG_AUDITSYSCALL #ifdef around loginuid | 09:06 |
| stekern | the strace also complains about missing /dev/log | 09:07 |
| blueCmd | stekern: did you see https://sourceware.org/ml/binutils/2014-04/msg00226.html ? | 10:55 |
| stekern | blueCmd: yes, didn't you see my \o/? | 11:28 |
| stekern | ah, no I mean ;) | 11:42 |
| stekern | are you sure that there wasn't any patches touching generic files that was left out? | 11:43 |
| stekern | I think I have seen some changelog in some generic place that ended with .or1k | 11:45 |
| stekern | I have atleast built our or1k-src with enable-cgen-maint recently | 11:55 |
| stekern | I can take a closer look in the evening | 11:56 |
| blueCmd | stekern: thanks | 12:19 |
| blueCmd | stekern: olofk jeremybennett: I got two requests for upstreaming GCC as well now today :) | 16:12 |
| stekern | blueCmd: https://github.com/openrisc/or1k-src/blob/or1k/cgen/ChangeLog.or1k | 17:01 |
| blueCmd | stekern: is the ChangeLog.or1k needed for cgen to generate stuff? | 17:03 |
| olofk | blueCmd: Yeah, it would be great to get the gcc ball rolling. Just need someone to dig through the SVN and git archives | 17:33 |
| stekern | blueCmd: not the changelog, but the changes it mention | 17:56 |
| blueCmd | olofk: yes.. we need.. someone... someone.. | 17:57 |
| blueCmd | stekern: maybe that wasn't included, hm | 17:57 |
| stekern | were those included in the patches? | 17:58 |
| blueCmd | stekern: no changed to cgen what I can see :( | 17:58 |
| blueCmd | changes* (in the patches) | 17:58 |
| blueCmd | http://openrisc.net/debian/patches/binutils-gdb/ those are the patches | 17:59 |
| stekern | ok, hopefully they are not controversal | 18:00 |
| blueCmd | stekern: mind doing a patch for that? otherwise I will do it | 18:12 |
| stekern | there are some generic changes to sim/ too | 18:43 |
| stekern | ah, but we didn't include the sim stuff at all? | 18:44 |
| stekern | maybe it's better if you put out the patch, you got a better idea of what's been sent and not | 18:45 |
| blueCmd | stekern: sure | 18:54 |
| blueCmd | stekern: I sent the things I used to build the debian package basically | 18:55 |
| blueCmd | I knew that gdb was broken, and I didn't want to send or1ksim | 18:55 |
| blueCmd | not sending cgen was a misstake, but what is sim used for? | 18:55 |
| stekern | yeah, or1ksim is a definite no | 19:02 |
| stekern | the sim is a nifty little thing, it's an autogenerated simulator from the cgen description | 19:03 |
| stekern | and it's commonly used to run the gcc regression suite | 19:05 |
| blueCmd | I see | 19:07 |
| blueCmd | stekern: diff -Nur . /srv/or1k-devel/or1k-src -x configure -x configure.ac -x sim -x gdb -x readline -x .git -x blt -x boards -x cgen -x contrib -x dejagnu -x expat -x expect -x itcl -x iwidgets -x libgloss -x libgui -x mmalloc -x newlib -x proto-toplev -x rda -x README-OR1K -x sid -x site.exp -x tcl -x tix -x tk -x utils -x winsup -x ChangeLog.or1k | 19:12 |
| blueCmd | that's how I made the patch | 19:12 |
| stekern | i see, I think the rest is ok but cgen and potentially sim | 19:18 |
| stekern | if you want, you can ping peter about it too | 19:19 |
| blueCmd | *sigh* It's a frikken pain to extract these patches | 19:20 |
| blueCmd | probably easier to just grab the diff on https://github.com/openrisc/or1k-src/commits/or1k/cgen | 19:21 |
| blueCmd | we should deprecate or1k-src, create or1k-binutils-gdb and make it a fork of the upstream | 19:23 |
| blueCmd | as I understand it, that was the way it was before - i think it makes the life of tracking the changes much easier | 19:23 |
| stekern | that's the way it is now, upstream, before it was unified in the humongous sourceware cvs repo | 19:25 |
| stekern | but yes, we should split out newlib and libgloss | 19:26 |
| blueCmd | stekern: do you have any idea where the cgen stuff is? I cannot find it in https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=tree | 19:42 |
| blueCmd | https://sourceware.org/cgi-bin/cvsweb.cgi/src/cgen/?cvsroot=cgen - oh | 19:43 |
| stekern | ah, so... those were rightfully omitted binutils patches I guess? | 19:45 |
| stekern | +from the | 19:45 |
| stekern | do cgen have an own ml? | 19:46 |
| stekern | ...I think it does | 19:46 |
| stekern | iirc juliusb sent mails to that with questions when he was working on this stuff | 19:46 |
| blueCmd | yes, I think I excluded it because the patch wasn't a patch | 19:46 |
| blueCmd | sim looks fine patch-wise though | 19:47 |
| stekern | the patch wasn't a patch? | 19:47 |
| blueCmd | hah, it makes sense! | 19:48 |
| blueCmd | the resulting patch of the diff wouldn't have been a patch but a copy | 19:48 |
| blueCmd | *sigh* that doesn't make sense either, hopefully you can puzzle something together | 19:49 |
| stekern | patch it together, you mean? | 19:54 |
| stekern | the loginuid was a red herring | 19:54 |
| blueCmd | when doing the diff between or1k-src and upstream I excluded stuff that was out-of-tree, which would have included gen | 19:54 |
| blueCmd | cgen* | 19:55 |
| blueCmd | stekern: oh? | 19:55 |
| stekern | ah, I see | 20:07 |
| stekern | yeah, I hack-enabled CONFIG_AUDITSYSCALL, which gives me the loginuid file | 20:09 |
| stekern | but the result is the same | 20:09 |
| stekern | however, if I do 'su -s /bin/dash stefan' I get a dash shell | 20:10 |
| stekern | ... and every command get served with a 'killed' message | 20:10 |
| stekern | it doesn't work if I change the shell in /etc/passwd | 20:11 |
| blueCmd | I wonder what happens if you run the "uid" binary I told you to run | 20:21 |
| blueCmd | as a shell or something | 20:21 |
| stekern | the one that printed 0 0 0? | 20:25 |
| blueCmd | yes | 20:26 |
| blueCmd | it should print 1000 1000 1000 for your user | 20:26 |
| stekern | yup, let's see what it does when I find it | 20:26 |
| blueCmd | stekern: http://c17826b0802a1966.paste.se/ that's cgen | 20:28 |
| stekern | so it is ;) | 20:31 |
| blueCmd | I guess I just post it to the very active mailinglist | 20:36 |
| stekern | I take that as irony | 20:37 |
| stekern | no luck running the uid binary... | 20:38 |
| stekern | was there a logsave in initramfs/bin before I did: sudo cp hacks/logsave initramfs/bin | 20:43 |
| stekern | ? | 20:43 |
| blueCmd | stekern: yes! | 20:45 |
| blueCmd | there was | 20:45 |
| blueCmd | you shouldn't do that | 20:45 |
| blueCmd | hah | 20:45 |
| blueCmd | that's an old instruction | 20:46 |
| stekern | yes, I noticed that you have removed that | 20:46 |
| blueCmd | https://github.com/bluecmd/or1k-debian/blob/master/README | 20:46 |
| blueCmd | ah good | 20:46 |
| blueCmd | reinstall e2fsprogs | 20:46 |
| blueCmd | or whatever the name is for that thing | 20:46 |
| blueCmd | e2fs* | 20:46 |
| stekern | but you pushed that after I had followed your old instruction | 20:46 |
| stekern | ...that I have already done, but there's no logsave in /bin | 20:47 |
| stekern | there is however one in /sbin/ | 20:47 |
| blueCmd | ah | 20:50 |
| blueCmd | just remove the dummy one then | 20:51 |
| stekern | that I have also done | 20:51 |
| stekern | so, I think I'm all good | 20:51 |
| stekern | is there any syslogd that we can use? | 20:53 |
| blueCmd | I'll see what I can do | 20:56 |
| stekern | because I guess su is trying to tell me something when it tries to open /dev/log | 20:58 |
| stekern | and I'd like to know what | 20:58 |
| blueCmd | touch /dev/log ? :) | 21:05 |
| stekern | mnja, it wants it to be a socket me thinks | 21:06 |
| blueCmd | stekern: https://sourceware.org/ml/cgen/2014-q2/msg00001.html | 21:34 |
| blueCmd | stekern: apt-get install rsyslog | 22:18 |
| --- Log closed Thu Apr 24 00:00:19 2014 | ||
Generated by irclog2html.py 2.15.2 by Marius Gedminas - find it at mg.pov.lt!