Re: [gentoo-catalyst] Too many problems with groupadd when creating stage3
On Thu, 2009-10-15 at 13:52 -0400, Mike Frysinger wrote: On Thursday 15 October 2009 13:50:40 Mike Frysinger wrote: or just fix catalyst erm, that's probably wrong. catalyst does what it's told as directed by the portage tree profiles. so the profiles probably need tweaking to produce a sane stage2. Tried it. If you'd prefer, I can pull up the list of bugs where we hit this on nearly every release, but since a viable solution was never reached, I don't see how it would help. Our main fix was to not build stages for desktop profiles, or to use a complete stage3 tarball as input for another stage3, but that's equally as broken. I'd *gladly* fix catalyst to work around this, if I had a clue how to do so. Hacking profiles ended up yielding nothing workable, just so everyone knows. -- Chris Gianelloni Developer wolf31o2.org signature.asc Description: This is a digitally signed message part
[gentoo-catalyst] ANNOUNCE: catalyst 2.0.6.901
I've released catalyst 2.0.6.901 for testing. The ebuild is available in my overlay. Simply layman -a wolf31o2 and you're good to go. I'll be honest, there's not been a ton of testing on this version. I'm going to be testing it tomorrow and fixing any major bugs that I find. Feel free to file bugs on Gentoo's bugzilla, under Gentoo Hosted Projects/catalyst, as I still get all the bug reports from there. -- Chris Gianelloni Developer wolf31o2.org signature.asc Description: This is a digitally signed message part
[gentoo-catalyst] ANNOUNCE: genkernel 3.4.10.901
Similar to the recent catalyst announcement, there's a new genkernel for testing. It's also available in my overlay, and bugs still go to Gentoo Hosted Projects/genkernel. Thanks, -- Chris Gianelloni Developer wolf31o2.org signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Autoresuming hides modifications in config file?
On Thu, 2008-09-18 at 08:52 -0500, Andrew Gaffney wrote: Amit Dor-Shifer wrote: Is that the way autoresume's intended to work? If so, should I clear autoresume points upon every modification in a conf file? Catalyst doesn't store any metadata about the spec file for a particular build, so it has no way to know if the spec has changed. That will probably chance in the future. For now, just make sure to use -a when you change the spec file. The general rule I use is to use -a if I changed anything that affects dependencies, or if I'm making a change in a section that has already been processed. The autoresume support was really designed to recover from failed merges, rather than to be a robust solution. As Andrew said, I'd eventually like to make the autoresume much smarter, which means we'd likely need to store the spec file in the autoresume data. If the keys in the spec for a particular function has changed and that function has already been run, delete the autoresume for that function. As we make things more and more modular, this could allow for an autoresume of a nearly-completed build where only the changes need to be processed. Using your rcadd as an example, let's say you were at the point of the ISO being made and realized that you forgot to add a service. You hit CTRL+C, edit the spec, then re-run catalyst, which would see that only the rcadd changed, so it would remove the autoresume data for the rc-update function (targets/livecd-stage2/livecd-stage2-controller.sh) and process only that function and the remaining ones, giving you an ISO with the least amount of processing time. -- Chris Gianelloni Developer wolf31o2.org signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] stage4 glibc ebuild stomps my locale.gen
On Thu, 2008-09-04 at 04:51 +0200, Peter Stuge wrote: Andrew Gaffney wrote: I think someone else should also have noticed this too. Yes, we've noticed. However, releng doesn't build stage4 tarballs, so we didn't care. I care. Is b.g.o the right way to go? Yes. which we obviously don't support :) Sure. It was a somewhat desperate attempt. A workaround might be to put a hacked glibc ebuild in my overlay, but I would like to avoid that upkeep in the long run. Well, it needs to be fixed in the glibc ebuild, rather than catalyst. File a bug against glibc and feel free to add [EMAIL PROTECTED] to the CC list, so we can watch/track the bug, also. -- Chris Gianelloni Developer wolf31o2.org signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Lions and Tigers and Forks, Oh My!
On Tue, 2008-07-29 at 04:16 +0200, lurker wrote: On 28/07/08 18:18, Chris Gianelloni wrote: Just a little FYI for everybody on the list... I've gone and forked most of the Release Engineering tools. I no longer plan on maintaining the Gentoo versions of catalyst, genkernel, hwdata, or livecd-tools. Instead, I am working on these projects under my own domain and will be releasing them as standalone applications, not under the Gentoo moniker. Now, these projects always have been Gentoo Hosted Projects and not really owned by the Gentoo project, but continued issues from other developers has proven that many don't understand this case. As such, it was simply easier to break away from the main project. This also allows us greater flexibility in making changes that would otherwise not be accepted within Gentoo. All of the current maintainers of these forks will be working on them in their new homes, making the Gentoo-hosted versions essentially dead. Interesting development. Can we be confident that the new independent catalyst will not break the current interfaces (ignoring trivial changes) and the use of portage? More to the point: what does this mean for existing catalyst projects that are using portage and intend to stay with it? There will be some disruptive changes to the spec interface. For one, I've removed all of the deprecated spec key support, so if your specs were based on catalyst 1.x specs, they won't work anymore (without modification.) As Andrew said, we don't intend on breaking anything, except where absolutely necessary. We're trying to focus on making catalyst easier to use, yet still more powerful. One example of an incompatible change that we're *definitely* going to adopt: compilation of bootloader. Since this obsoletes the cdtar key, that option won't be valid in newer catalyst versions. This will be the general scope of interface-level changes, so people relying on catalyst, such as myself, have no need to worry. Things will stay similar-enough that transition from current catalyst to the new one will be rather minor. -- Chris Gianelloni Release Engineering Strategic Lead Games Developer signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] small distros - setups and practices?
On Mon, 2008-05-05 at 20:02 +0200, Robert Šmol wrote: Can you explain more about your project? Not really... it's something under development and quite a ways away from seeing the light of day. PORTAGE_BINHOST=http://packages.builder.domain/machine-type/ Well, I'm using catalyst to do the building. Slave - hey, this is my /var/lib/portage/world, my USE flags (/etc/make.conf) and I want such and such package Wow! That sure sounds a lot like what catalyst does. If only we had a way to specify settings for packages and such... something like a profile. Maybe even if we had a file we could give catalyst, some kind of specification... Oh wait... *grin* Master - calculating a bit and providing Slave with packages catalyst -f packages.spec rsync --archive --delete-after /var/tmp/catalyst/packages/$rel_type/$target-$subarch-$rel_ver/* master:/exports/packages/$target/$subarch/$rel_ver (but this is overkill) ...or is it? -- Chris Gianelloni Release Engineering Strategic Lead Games Developer signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] small distros - setups and practices?
On Mon, 2008-05-05 at 21:41 +0200, Ramon van Alteren wrote: Is there any public repository where you are developing this stuff in ? Nope. curious why you'd want to re-implement either puppet of cf-engine. I never said anything about re-implementing anything. I plan on leveraging existing technologies, where possible. -- Chris Gianelloni Release Engineering Strategic Lead Games Developer signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] small distros - setups and practices?
On Fri, 2008-05-02 at 10:20 +0800, Max Arnold wrote: Sure. In fact, I have 2 different host types right now, a master and a build host though they can reside on the same machine. Which roles they have? I'm not sure that I follow. Are you asking what specific tasks would be performed by the roles above? Yes, what specific tasks would be performed by these host types and what are benefits of such distinction? I guess a master purpose is to serve updates, or things are more tricky? Yeah, you're thinking way too small. ;] The master server will be the control interface for custom profiles and binary repositories, which are built by the build servers. It will (eventually) also be the configuration engine for each server. I hope to eventually have a complete system capable of controlling a Gentoo system from birth to death, from a centralized location. -- Chris Gianelloni Release Engineering Strategic Lead Games Developer signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] small distros - setups and practices?
On Wed, 2008-04-23 at 10:03 +0800, Max Arnold wrote: I want special build host, where all binary targets are built. Its setup should be easily repeatable by other developers (i.e. every task should be scripted, without manual tweaks like 'chroot here and patch/emerge this'). Catalyst with specs and overlays seems fine to me. Sure. In fact, I have 2 different host types right now, a master and a build host though they can reside on the same machine. Which roles they have? I'm not sure that I follow. Are you asking what specific tasks would be performed by the roles above? -- Chris Gianelloni Release Engineering Strategic Lead Games Developer signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] proper way to deviate in system use flags?
On Sat, 2008-04-19 at 11:19 +0800, Max Arnold wrote: On Fri, Apr 18, 2008 at 03:12:08PM -0700, Chris Gianelloni wrote: My steps are: stage3-i686-2007.0 + recent snapshot - stage1 - stage2 - stage3 - stage4 OK, start from an x86 stage3, though. Is it critical? I'm shipping only stage4 and it will be deployed on fixed hardware configuration. If you're building a stage1, you should always use a generic stage3 as your seed. It likely isn't critical for you, but it's good practice to get used to doing. So it is seems that my stage4/use and portage_confdir does not affect system packages (I guess that catalyst does only --emptytree when emerging stage4). Ehh, not quite. The USE you set in your spec only affects the stage build. It doesn't affect what happens *after* the stage build. It also doesn't affect packages built before your current stage build. One thing that you did not mention is what version of catalyst that you're using. Newer versions of catalyst, like the currently masked 2.0.6_pre17, support writing out spec_prefix/use to make.conf so the changes stick in your stage. I'm using catalyst-2.0.5, but at least on stage4 it puts my stage4/use into make.conf and copies content of portage_confdir to /etc/portage. I've just verified it in resulting stage4 tarball. Sure, but none of your earlier stages did this correctly and even the stage4 target doesn't get it quite right. Upgrade your catalyst version. So, there are my questions: 1. Am I correct in assumption that use flags are stacked during stage4 like this: profile - stage4/use - package.use (increasing priority from left to right)? That's over-simplified, but yes. The reason I'm asking this is a phrase If you set the portage configuration dir: package.use doesn't work how you think it should in catalyst. It's a known bug, but one that won't likely be fixed for some time. Basically, if you use $target/use, then it will ignore package.use settings. here: http://gentoo-wiki.com/HOWTO_build_a_LiveCD Is that already fixed? It should be *only* in newer catalyst, not the current stable. 2. What is the proper (and simple to maintain) way to deviate slightly in system use flags? The way that you're doing it works fine if you're planning on shipping a stage4 all the time. Were I doing this myself, I'd make a new profile with the desired changes, but that's just me. My guesses: 1. Add 'hostuse' variable to earlier specs (stage3?) Already done in newer catalyst versions... 2.0.5 uses it: generic_stage_target.py adds hostuse to valid_values list, and stage4_chroot.sh mentions $clst_HOSTUSE. But is it safe to simply define hostuse in stage spec? Is it additive (I see x86.py assigns some values to this variable) or arch specific value will be simply overridden and it is better to do not redefine it? You cannot redefine it. It is defined exactly once. HOSTUSE is literally *only* for subarchitecture-specific USE which controls CPU capabilities/features. It is only designed to be used internally, so if you try to do anything with it, I cannot guarantee what will happen. 4. Create my own profile (don't know where to put it and how to maintain during tree updates) This is what you should do. Simply stick it in $portdir/profiles where you'd like it before running your snapshot. To keep emerge --sync from wiping it out, add PORTAGE_RSYNC_EXTRA_OPTS=--exclude=profiles/myprofilename to your /etc/make.conf on your host/build system. Can be custom profile added to portage_overlay on all stages istead of putting it to tree? Yes, it can, but it is safer to put it into your tree that you're using for your snapshots. I suggest keeping a separate portage tree from /usr/portage for this, so you don't have to worry about accidentally hosing it. BTW, what is the difference between profiles/default/linux and profiles/default-linux ? Well, profiles/default/linux is the new location and default-linux will be going away, so that makes it rather simple. ;] So for now I'll try to upgrade catalyst, add portage_confdir to all stages, leave only stage4/use (for additional world packages) and experiment with profiles. Sounds great! I also have some more general questions, but I'll ask them in another thread. OK. Thank you! -- Chris Gianelloni Release Engineering Strategic Lead Games Developer signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] livecd-stage1 problem
On Thu, 2008-03-13 at 18:23 -0600, Daniel Díaz wrote: On Thu, Mar 13, 2008 at 11:22 AM, Chris Gianelloni [EMAIL PROTECTED] wrote: [...] I plan on starting up a fairly comprehensive manual for catalyst after this coming release. Chris, if I can be of any help please let me know . Somebody else might be interested too in collaborating developing a good reference manual. Absolutely, I plan on starting it simply as a listing of every spec file key available in each target, but from there, plan on writing up a simple howto which will slowly expand into a comprehensive manual. I would definitely be interested in help. Do you know GuideXML? -- Chris Gianelloni Release Engineering Strategic Lead Games Developer signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] livecd-stage1 problem
On Tue, 2008-03-11 at 22:33 -0600, Daniel Díaz wrote: Should I update that howto? It recommends getting the stage3 off a mirror. I just noticed the Wiki page has been updated. Thanks many! I'm sure many have been bitten by that bug, and even if that document out of this project's scope, surely that'll keep people that found it somehow useful (like you) from falling for that. I updated it so that hopefully people will quit hitting this bug when doing their own builds. -- Chris Gianelloni Release Engineering Strategic Lead Games Developer signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] livecd-stage1 problem
On Wed, 2008-03-12 at 11:14 -0400, Robert C Corsaro wrote: Also, the wiki is *not* an official source of Gentoo . I wish people would realize that. I think that the problem is, there is no official documentation AFAIK. Yes, there is. Try to emerge catalyst again and read what it actually says. There is no *web* documentation. The catalyst page points to the newbie wiki as the best source of documentation. It was the only path I could find to understanding how to use catalyst. Saying it's not official documentation is fine, but it Umm... no. The catalyst page most definitely does *not* point to *anything* on gentoo-wiki. I'm guessing you mean some random catalyst page on gentoo-wiki, don't you? will still help reduce the amount of questions asked here to try and make it 'right'(which I have no problem doing). I'm sorry if the information I was looking for is written down somewhere and I didn't find it. If it is, perhaps we can add that docs location to the catalyst page. I did search the mailing list archive and didn't see my problem. Maybe I didn't look hard enough? If not, sorry for wasting time. I know it sucks to answer the same question over and over. Every livecd-stage1 blocker issue ever reported has been this same issue. I know it's hit this list a few times, but would likely be difficult to search and find, since I bet most people would search for the package in question, rather than the action being taken. BTW, catalyst is awesome, so thanks to everyone who worked on it. Well, check out AUTHORS and send them beer. ;] -- Chris Gianelloni Release Engineering Strategic Lead Games Developer signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] livecd-stage1 problem
On Wed, 2008-03-12 at 11:25 -0400, Robert C Corsaro wrote: What is the recommended way to learn to use catalyst? # emerge catalyst # mkdir -p /release # cd /release # svn co svn://svn.gentoo.org/releng # cd releng/trunk/releases/2007.0/specs/amd64 # sed -i 's:/root/livecd/kconfig/2007.0/amd64:../../kconfig/amd64:' *.spec # mkdir -p /var/tmp/catalyst/snapshots # cd /var/tmp/catalyst/snapshots # wget http://gentoo.osuosl.org/releases/snapshots/portage-2007.0.tar.bz2 # mkdir -p /var/tmp/catalyst/builds/default # cd /var/tmp/catalyst/builds/default # wget http://gentoo.osuosl.org/releases/amd64/2006.1/stages/stage3-amd64-2006.1.tar.bz2 Now, at this point, you'll have a proper checkout and be ready to go. I don't recommend using the 2008.0 specs until after they're finalized and the release is made. It's very simple. Start with stage1, then stage2, then stage3, then proceed on to your different CD media. I'd suggest starting with installcd-stage1, rather than livecd-stage1. Make changes. See what those changes actually end up doing. Read the specs in /usr/share/doc/catalyst-*/examples... -- Chris Gianelloni Release Engineering Strategic Lead Games Developer signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] livecd-stage1 problem
On Thu, 2008-03-13 at 13:16 -0400, Robert C Corsaro wrote: Well, check out AUTHORS and send them beer. ;] Send beer?! Anyone live in Boston? ;) Dammit. I knew I should have taken the job in Boston and not Mountain View. ;] I plan on starting up a fairly comprehensive manual for catalyst after this coming release. -- Chris Gianelloni Release Engineering Strategic Lead Games Developer signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] livecd-stage1 problem
On Tue, 2008-03-11 at 15:45 -0400, Robert C Corsaro wrote: dev-perl/XML-Parser is definitely installed. I'm using the release/current stage3 archive for this build. I also tried the 2007.0 release with no luck. Perhaps I need a newer stage3? Well, there's your problem. To try to keep people from screwing this up even though we've said it repeatedly, I edited the wiki page. Also, the wiki is *not* an official source of Gentoo . I wish people would realize that. -- Chris Gianelloni Release Engineering Strategic Lead Games Developer signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Catalyst 2.0.6_pre9 mksquashfs hangs building squashfs + solution
On Mon, 2008-02-25 at 18:05 -0600, Andrew Gaffney wrote: Chris Gianelloni wrote: On Mon, 2008-02-25 at 22:58 +0100, Pongracz Istvan wrote: Probably the situation is not the same, I did not compared it to my version (3.3). This also happens if you're using mksquashfs 3.3 and trying to make a 3.1 squashfs, like would be done if your kernel is 2.6.23 and you're using ~arch on your build machine. We *want* the sparse option, so the fix is to use the right version of mksquashfs for the kernel you're building/deploying. Eh, he was talking about mksquashfs hanging while creating the squashfs image, not about the kernel freaking out while trying to mount it. I know exactly what he was talking about. -- Chris Gianelloni Release Engineering Strategic Lead Games Developer -- gentoo-catalyst@lists.gentoo.org mailing list
Re: [gentoo-catalyst] mkxf86config: Virtualbox experience
On Tue, 2008-02-26 at 11:24 +0100, Pongracz Istvan wrote: Is that possible to make the default driver to vesa, instead of ? This is a discussion forum. If you have feature requests, file a bug report. And I have no patch for this, sorry. Well, patches are about the only way you're going to get many of these things changed, especially since I sometimes have a hard time following what you're asking. Patches are simple and are quite explicit about what changes you want/intend. -- Chris Gianelloni Release Engineering Strategic Lead Games Developer -- gentoo-catalyst@lists.gentoo.org mailing list
Re: R: [gentoo-catalyst] producing a DVD
On Tue, 2008-02-26 at 14:43 +0100, Lanza Fabrizio wrote: That is exactly what I cannot do. I insert a DVD in the CD/DVD drive. I start Nero, choose burn image, I select my .iso, and it ejects the DVD Remember that we're burning using Linux-based tools, not Windows-based ones. I have no clue how to provide you support for Windows, nor will I even try. Sorry. Try burning it under an OS that doesn't suck and if you're still having problems, let us know. Every CD/DVD mastering software will check what you are going to burn. If it has a CD project/filesystem, it won't burn it on a DVD. Umm... no. A DVD can be *either* UDF or iso9660. If your software is refusing to work with 700MB ISO9660 images, it is broken. Also, since qemu does not load CDs larger than 700M... there is a problem. Broken. File a bug. I also noticed a curious thing. When I mount the .iso (larger than 700M) with a program such as Daemon Tools, it won't show the filesystem. It shows as it is completely empty. Windows. Try the same under Linux and it'll work. I would rather find out the way I can get a DVD iso with Catalyst. Well, catalyst produces images which are within the ISO spec. If your tools do not work with them, submit patches. Otherwise, we're not going to change anything, as we're within the defined specification which other tools are supposed to follow. -- Chris Gianelloni Release Engineering Strategic Lead Games Developer -- gentoo-catalyst@lists.gentoo.org mailing list
Re: R: [gentoo-catalyst] producing a DVD
On Tue, 2008-02-26 at 08:36 -0600, Andrew Gaffney wrote: I would rather find out the way I can get a DVD iso with Catalyst. I already told you. Catalyst doesn't do anything special for DVDs, and the larger ISO images it generates work just fine. There is zero support for creating UDF images in catalyst. Use another program to burn the image. Nero obviously has an issue. Burning a larger ISO with cdrecord works without a problem. I use: cdrecord -v -sao dev=/dev/sr0 driveropts=burnfree /path/to/iso I've been able to successfully burn the 4GB LiveDVD for both AMD64/x86 and had exactly 0 issues in about 40 burns. Again, your tools are being stupid. They are thinking that the image *must* be UDF when they are not required to be, at all. -- Chris Gianelloni Release Engineering Strategic Lead Games Developer -- gentoo-catalyst@lists.gentoo.org mailing list
Re: [gentoo-catalyst] Catalyst 2.0.6_pre9 mksquashfs hangs building squashfs + solution
On Mon, 2008-02-25 at 22:58 +0100, Pongracz Istvan wrote: Probably the situation is not the same, I did not compared it to my version (3.3). This also happens if you're using mksquashfs 3.3 and trying to make a 3.1 squashfs, like would be done if your kernel is 2.6.23 and you're using ~arch on your build machine. We *want* the sparse option, so the fix is to use the right version of mksquashfs for the kernel you're building/deploying. -- Chris Gianelloni Release Engineering Strategic Lead Games Developer -- gentoo-catalyst@lists.gentoo.org mailing list
Re: [gentoo-catalyst] Build stage4 ssmtp conflicts with postfix.
On Fri, 2008-02-22 at 11:34 -0600, Andrew Gaffney wrote: I am building stage4, but he return error where ssmtp conflicts with postfix. I trying configure virtuals in my profile to set postfix virtual mta, but not work. Somebody can help me? This is a portage bug and has nothing to do with catalyst. You'd encounter this same problem when emerging the same packages in a stage3 chroot. No, it is a portage feature. You cannot install two packages which block each other. If you want postfix in your stage4, build a set of lower stages (you should *always* do this, anyway) with your customizations (like virtual/mta being postfix) and then rebuild your stage4. -- Chris Gianelloni Release Engineering Strategic Lead Games Developer signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] reference manual gone?
On Thu, 2008-02-14 at 21:06 -0600, Andrew Gaffney wrote: Speaking of example specs, are the ones for 2008.0 available publicly yet? Yes, but they're still a work in progress, which they will be right up until we actually release. They're available from the shiny new releng repo in gentoo's SVN. Indeed. I wouldn't use them. At this point, most arches don't even build completely. Also, you don't have the 2008.0 snapshot (and you won't get it... :P) so the specs are useless, since we switched to a new multi-parent profile layout. In fact, it is this layout which is causing most of our problems. If you want *working* specs, grab 2007.0 (or even 2007.1) and work from there. Using the in-progress specs is *never* a good idea, since they could be broken, especially since most release coordinators don't update their specs in SVN until after things have built properly. -- Chris Gianelloni Release Engineering Strategic Lead Games Developer signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] fooling uname
On Wed, 2008-02-13 at 22:11 -0500, Mike Frysinger wrote: On Wednesday 13 February 2008, lurker wrote: Currently catalyst (2.0.5, 2.0.6pre4) doesn't seem to change what uname outputs. uname -r, for example, prints the version of the kernel running catalyst, not the kernel that's in the stage. Of course, that's how it is, but isn't it possible to hack that somehow? It's a bit tedioius to patch everything that depends on uname (e.g. the raling-rt61 ebuild). Can the uname string be set by some envirnoment variable in an env script specified in the spec file? stages dont have kernels Indeed. As Andrew has said and Mike as alluded to, not a single catalyst target has a kernel compiled prior to stage4/livecd-stage2, so there's no reason for this. Any ebuild that uses uname for *anything* is broken and a bug should be filed to have the package fixed. -- Chris Gianelloni Release Engineering Strategic Lead Games Developer signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] catalyst related kernel problem?
On Sat, 2008-01-12 at 02:20 +0100, lurker wrote: In my case, the only way I could get it to work consistently was to patch genkernel to use symbolic links instead, when setting up the initramfs. After upgrading to busybox-1.8.2 I've not experienced this any more. Perhaps it's just a coincidence due to the random nature of this problem. Are you talking about the host system or in your genkernel.conf ? -- Chris Gianelloni Release Engineering Strategic Lead Games Developer signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] catalyst related kernel problem?
On Sat, 2008-01-12 at 18:15 +0100, lurker wrote: On 12/01/08 17:49, Chris Gianelloni wrote: On Sat, 2008-01-12 at 02:20 +0100, lurker wrote: In my case, the only way I could get it to work consistently was to patch genkernel to use symbolic links instead, when setting up the initramfs. After upgrading to busybox-1.8.2 I've not experienced this any more. Perhaps it's just a coincidence due to the random nature of this problem. Are you talking about the host system or in your genkernel.conf ? What I did was simply to add busybox-1.8.2 to the stage3 build, then rebuild the livecd stages and generate the iso. But now I realise that it probably doesn't have anything to do with what genkernel do later on (?) as genkernel seems to use its own busybox (version 1.1.3) which is uses for its initrd images. So I guess I just was lucky... Well, sometimes luck counts... =] -- Chris Gianelloni Release Engineering Strategic Lead Games Developer signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] livecd stage1 kernel .config
On Thu, 2008-01-10 at 12:19 +0100, Philipp Riegger wrote: On Mon, 2008-01-07 at 14:33 -0500, Greg Fitzgerald wrote: It compiles around 400 packages then dies on gnome-menus which requires a .config file in my kernel sources directory. Wondering how to solve this issue. I am new to making a livecd with catalyst, but my understanding is you set the .config file for your kernel in your livecd stage2. I tried to copy it to the livecd cache directory under /usr/src/linux and starting again with -a but it appears to get deleted when I start again. Can anyone let me know how I should proceed from here? I have the same problem when compiling gnome with tinderbox. What is the right way to solve this one? Get the package updated to not need to do a stupid kernel configuration check is the proper solution. Quite honestly, the linux-info eclass needs to be completely rewritten. The eclass should *not* fail by default, for any reason. Any failures should be optional and opt-in only. The linux-mod eclass, on the other hand, *should* fail. Basically, not being able to get information about the running/compiled kernel in a package which is not a kernel module should never be a failure. In the current code, it is a failure. http://bugs.gentoo.org/show_bug.cgi?id=192556 Everybody who is getting this error, add yourself to CC on that bug. That bug being fixed is what you need to resolve this issue. -- Chris Gianelloni Release Engineering Strategic Lead Games Developer signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] LiveCD fs root creation
On Mon, 2007-11-19 at 13:06 +0100, lurker wrote: How/when/where is the LiveCD's filesystem root created? With this I'm referring to the actual fs root (/) when running the CD, not the root found in the squashfs image that is created during build time. For some reason, these are not the same. It appears that only some directories from the image appears on the LiveCD root, where some symlinks not present in the image root also get created. A typical squashfs-enabled LiveCD built with catalyst has 3 different places that could be considered root on the CD at different times. The first is the outside root, what you see when you mount the CD and browse it. The second is the intramfs root which houses the splash themes, drivers, busybox, lvm/evms, etc. from genkernel. The last, is the tmpfs created at boot, which genkernel runs one of switch_root, pivot_root, or chroot. This tmpfs consists of certain directories from within the squashfs image, like /etc and /tmp, so they're writable. It also symlinks certain other directories, like /usr. In my case I need /media to have KDED media manager automounting through HAL. I explicitly create /media in my fsscript and have verified that it indeed is created (by mounting the image) but it still doesn't appear on the LiveCD root. This is also the case for the Gentoo LiveCD's. The obvious solution is to create /media in local.start at every LiveCD boot up but I would nonetheless like to know the procedure of how the LiveCD root is set up from the contents of the image and if and how I can modify it. You could modify it, but you would need to modify genkernel to do so. There are two ways that I can think of to resolve this. Using local.start is one way of doing it, and how we have done some things in the past. We're trying to move away from that in the official media, but it's still a viable solution. The second is to use a feature of genkernel that very few people know exists. When booting a CD, genkernel will search for ${NEW_ROOT}/mnt/cdrom/cdupdate.sh, which is the outside root of the CD. When genkernel finds this, it executes it. This is done in the second root above, the initramfs, immediately prior to the chroot/pivot_root/switch_root. Since it runs in busybox, it needs to run in the minimal busybox shell correctly. In your case, it could simply say: #!/bin/sh mkdir -p ${NEW_ROOT}/media Make sure the file is executable, or genkernel won't run it. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Building stage3 fails because of no C++ compiler
On Fri, 2007-11-16 at 08:51 +0100, Jan Stępień wrote: I've been trying to create a live CD using catalyst building my own stage(1-3). I've downloaded stage3-i686-2007.0.tar.bz2 (MD5 sum 55f18797e55a19f1f7d43d649f416504) and ran: Use a generic stage. Do not use an i686 stage3 as a seed, use an x86 one. Your stage1 subarch should be x86 not i686... I would also suggest not using the C(XX)FLAGS you've chosen. Add a -march=i686. You're already building i686-only binaries via subarch i686, it makes no sense to under-optimize the binaries, too. ;] Let us know if those things didn't fix it. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Building stage3 fails because of no C++ compiler
On Fri, 2007-11-16 at 23:54 +0100, Jan Stępień wrote: What am I doing wrong? Should I change CHOST only between stage2 and 3? Or is there a different solution? I'd be grateful for your support. .spec files attached. You should change the CHOST in the stage2 spec. Look at http://sources.gentoo.org/viewcvs.py/gentoo/src/releng/specs/2007.1/x86/ for examples. Also, there's no need to post your specs every time. It just adds a bunch of stuff for people to download that don't necessarily need it. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] howto add packages to livecd .specs?
On Thu, 2007-10-25 at 13:27 -1000, Joshua Hoblitt wrote: I'd like to add rsync to the 2007.0 livecds specs but I'm not having much luck. I added net-misc/rsync to installcd-stage1-minimal.spec:livecd/packages and reran `catalyst -p -f installcd-stage1-minimal.spec catalyst -p -f installcd-stage2-minimal.spec` but rsync didn't appear in the resulting image. Am I missing a step or is the rsync binary being deleted in the stage2 build process? Well, rsync is in every stage, from stage1 on up. You're likely removing it in livecd-stage2. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Pre- and post- kmerge mismatch in generic-stage-target.py
On Mon, 2007-10-08 at 09:10 -0400, John Howard wrote: cause problems with more than one. Or I might be wrong in thinking that the pre- and post-kmerge steps should match up; if so please enlighten me. You're wrong. The pre-merge stage is only needed once. The kmerge and post-kmerge is needed for each kernel. Of course, it might be that over time things have changed enough that we no longer need to run the post-kmerge after each run, but we'd need to determine if that's the case or not. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] no stage1 created
On Fri, 2007-10-05 at 15:03 +0200, Robert Šmol wrote: OK, I was reading official docs yesterday and it seems there is only reference guide, but no high level view on how things works together. The reference guide is for catalyst 1.x and was never updated for catalyst 2.x, don't use it. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] no stage1 created
On Fri, 2007-10-05 at 17:36 +0200, Robert Šmol wrote: I still have some questions (like how to do my own stage3), but I'll dig through archives and play with catalyst for now. You take a stage3 tarball as a seed, then build stages 1, 2, then 3, in order. One thing to realize is the stages are not very customizable, so if you're not compiling for a specific architecture/subarch, the only way to really customize a stage is by modifying the profile which it uses. If you want completely customizable stages, you want a stage4. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Missing /root
On Wed, 2007-10-03 at 07:00 -0500, Andrew Gaffney wrote: When ccache is unmerged at the end of the stage build, it removes /root, since its the sole owner. Does that seem plausible? That seems very plausible, as Andrew said. However, that isn't how portage works. The only way that ccache could remove /root is if nothing else put any files there. Now, that might be the case here, which is why I still agree it to be very plausible. Adding a .keep into /root should resolve it. Fell like testing? Throw in a keepdir /root into the ebuild where the directory is created and see if it fixes it. It may even be the reason that wolf31o2 recommends that it not be used. Actually, I don't recommend it because I want everything to come from a complete binpkg or from source. I don't like having object files sitting around that might change between versions. I just simply don't trust ccache enough to consider it high-enough quality to be used on our releases. That being said, I could be completely wrong and it could be just fine. I'm sure it really is, but why take the chance? -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Files inside /home/user
On Wed, 2007-09-12 at 07:04 -0500, Andrew Gaffney wrote: Don't put the data in /home. You can also look into figuring out the (now broken, again) unionfs support in genkernel. I'll try to use the unionfs support; is there a genkernel version that you know support it? You missed his point. The latest gk versions have the unionfs support, but it keeps breaking for some reason. He was dropping a subtle hint for you to take a look at it and try to fix it ;) Correct. No current version of genkernel works with all of the relevant kernels. I was merely suggesting that there is a big bunch of code already that does what you want, but it isn't very important to the developers because we never got it working quite right, and as such don't use it, but it could be fixed. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Files inside /home/user
On Tue, 2007-09-11 at 16:38 +0200, Luca Casagrande wrote: Hello everybody. In my LiveCD, I have got about 100 MB of data inside the the user folder; if i boot the LiveCD on machine with 512 Mb of ram, I have got a lof of cp error during the Copying Read-Write to tmpfs phase. I think that the reason of this trouble is that /home/ is copied inside the tmpfs during the boot and it's being fulled with all my data. Is there a way for solving this? Don't put the data in /home. You can also look into figuring out the (now broken, again) unionfs support in genkernel. There are several places that are copied to tmpfs for proper functionality: /etc /home /root /tmp /var -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Can't start catalyst -f tinderbox.spec
On Thu, 2007-09-06 at 11:32 -0300, Daniel van Ham Colchete wrote: pkgcache_path: kerncache_path: Why are you doing these? Considering that catalyst spit an error on pkgcache_path, I'd say that's your problem. (subarch: nocona is right, I changed the arch/x86.py file to allow it) I'm sure it is simple, but please submit this patch as a bug report and I'll get it into the next catalyst release. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Can't start catalyst -f tinderbox.spec
On Thu, 2007-09-06 at 09:53 -0500, Andrew Gaffney wrote: Daniel van Ham Colchete wrote: Hello yall, after using catalyst to successfully build a stage4 for my servers, I'm now trying to play a little bit with tinderbox. I don't think anybody has played with catalyst's tinderbox support in a while. I use it every day. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] libtool dependencies in stage3
On Tue, 2007-08-28 at 10:51 -0400, Mike Frysinger wrote: On Tuesday 28 August 2007, Åsmund Grammeltvedt wrote: On Tuesday 28 August 2007 16:32:39 Mike Frysinger wrote: On Tuesday 28 August 2007, Åsmund Grammeltvedt wrote: Still, is it impossible for features to carry over from packages in the bootstrap stage and affecting depending packages before the dependency is reemerged, without this being a bug in the depending package? speaking generally, probably ... but here, it is a bug in attr, pure and simple ... it should not be using the host libtool Ok, I'm mostly trying to justify messing with /var/db/pkg to myself. Cleaning it out and bootstrapping from there seems like it would reduce the impact of such bugs. then you'd hit things like portage going crazy over package collisions ... or in the scenario where the older stage had an older version and the newer version has different files so you're left with orphaned cruft ... Correct. There is a reason that we moved *away* from removing the /var/db/pkg stuff for stage1. This is it. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Python 2.5 Livecd
On Tue, 2007-08-28 at 12:02 +0200, Luca Casagrande wrote: Hi to all! A friend of mine has asked me for a livecd that will got only python 2.5. I thought about using my old (with 2.4 python stable) stage3, and unmasking python during the build of livecd-stage1..but i don't feel good with this way. Do you think it's better rebuild the whole stages with a snapshot that use python 2.5? It definitely is better, since you could otherwise end up with artifacts from the older python versions and things might not work correctly. Pretty much if anything major changes in system (gcc, glibc, binutils, portage, python) it is usually a good idea to rebuild the stages. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] changing inittab
On Thu, 2007-08-02 at 10:28 -1000, Joshua Hoblitt wrote: rhetorical Why don't people trust that the guy who writes all this crap knows what he's talking about? /rhetorical :P Your assertion is that it's completely obvous that a variable named bootargs, that passes options to the kernel, would have other magical side effects? No, my assertion is that when I provide a solution to someone's query that I might know what I'm talking about. ;] Normally, yes, it would only be good for kernel output, but as someone else mentioned, you have livecd-functions.sh that re-writes the inittab. Well, guess what triggers it? That's right, a console= option on the command line. If you type (or via catalyst, add) console=something, livecd-functions.sh (via functions.sh in baselayout) will interpret it and modify inittab accordingly before we ever even hit runlevel 3, so it starts a console on that serial line for login. Seriously, can you cut and paste that paragraph into the manual? Ehh... What manual? There's no manual for catalyst that is current. Also, I don't have the time (or the desire) to document all of the idiosyncrasies of the packages that make up a Gentoo LiveCD build. We use quite a few packages that other people don't need to use on their own media. Remember that everything written by Release Engineering is designed for Release Engineering consumption. No guarantees are made to anyone else on anything. We guarantee that it will build a release the way we want it given a particular package set with specific versions, the exact ones used by us for our builds. Also, realize that livecd-tools is *not* catalyst and shouldn't be documented with catalyst. Thanks for your help. You're quite welcome. This list is pretty much here just so people can get help using catalyst, so feel free to ask anything that you need. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] changing inittab
On Wed, 2007-08-01 at 10:00 -0400, Scott Biddlestone wrote: If you installed livecd-tools, the inittab is overwritten by /sbin/livecd-functions.sh when the cd boots ( it is backed up to inittab.old ). One way to fix it is to modify the script by creating a new livecd-tools ebuild in your portage overlay - copy the current ebuild and just add the patch to it. Please don't touch livecd-tools. Don't touch sysvinit. Instead, do the following in your livecd-stage2.spec file... livecd/bootargs: console=ttyS0,115200 That's it. When the CD is built, it'll have the console appended to both the normal and nofb kernels. Either kernel will work perfectly fine for serial console, so you're done. -Scott Nelson Batalha wrote: I believe you can overwrite it in the overlay or you can do equery belongs /etc/inittab in your own system to discover it belongs to sys-apps/sysvinit, so you can change that too. Joshua Hoblitt escreveu: Hi Folks, I'm attempting to generate an amd64 livecd that starts up a tty on a serial port. So I want to do is add the line s0:12345:respawn:/sbin/agetty 115200 ttyS0 vt100 to the /etc/inittab. The trouble is, I can't figure out exactly where iniitab is coming from. Where does catalyst get /etc/inittab from? Is it direct from the stage1 seed? How should I go about modifying it? Thanks, -J -- -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] changing inittab
On Wed, 2007-08-01 at 11:55 -1000, Joshua Hoblitt wrote: On Wed, Aug 01, 2007 at 09:58:36AM -0700, Chris Gianelloni wrote: On Wed, 2007-08-01 at 10:00 -0400, Scott Biddlestone wrote: If you installed livecd-tools, the inittab is overwritten by /sbin/livecd-functions.sh when the cd boots ( it is backed up to inittab.old ). One way to fix it is to modify the script by creating a new livecd-tools ebuild in your portage overlay - copy the current ebuild and just add the patch to it. Please don't touch livecd-tools. Don't touch sysvinit. Instead, do the following in your livecd-stage2.spec file... livecd/bootargs: console=ttyS0,115200 That's it. When the CD is built, it'll have the console appended to both the normal and nofb kernels. Either kernel will work perfectly fine for serial console, so you're done. Eh, isn't that only good for kernel output? I'd like to actually be able to get a shell on the serial line... rhetorical Why don't people trust that the guy who writes all this crap knows what he's talking about? /rhetorical :P Normally, yes, it would only be good for kernel output, but as someone else mentioned, you have livecd-functions.sh that re-writes the inittab. Well, guess what triggers it? That's right, a console= option on the command line. If you type (or via catalyst, add) console=something, livecd-functions.sh (via functions.sh in baselayout) will interpret it and modify inittab accordingly before we ever even hit runlevel 3, so it starts a console on that serial line for login. Try it. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] ~arch builds
On Thu, 2007-07-19 at 12:18 +0300, Philipp Riegger wrote: Hi! I'm just curious (and did not find anything about that yet). Is it possible to build ~arch stages and livecds without changing the tree or using portage_confdir? It is possible, but not something we would ever support, so I couldn't even tell you if it would work. It *should* work but that doesn't mean that it *will* work. You'd just need to set ACCEPT_KEYWORDS in your envscript, though you have to realize that it won't affect make.conf in the stages, so the first time someone tried a world update, it would try to downgrade every package and likely break the person's system. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Error trying to build uclibc stages
On Thu, 2007-07-19 at 12:21 +0300, Philipp Riegger wrote: I also talked to agaffney on IRC and it seems that without any hacks like this it is not possible to build uclibc stages from a glibc seed, and with hacks like this it is not guaranteed to work lateron (the last half sentence was not discussed with him, but what i experienced). Pretty much. Your best bet is to build the uclibc stages, then use your uclibc stage3 as a seed and build another set of uclibc stages from it. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] changing genkernel's configuration files as invoked by Catalyst
On Wed, 2007-07-18 at 14:53 -1000, Joshua Hoblitt wrote: Hi Folks, I'd like add the arcmsr driver to the set of SCSI modules included in the initrd image generated by genkernel when run by catalyst. I understand what genkernel files need to be modified to accomplish this but I'm not sure how to do this under catalyst. Do I need to provide a portage overlay to catalyst or is there perhaps some magic variable that can be set to point to genkernels config files? You'll need to update genkernel, create a new tarball/ebuild, and resnapshot/portage_overlay it. Yeah, it sucks, but genkernel has no mechanism for choosing the drivers. Also, if this driver should be done by default, please file a bug and I'll fix it in genkernel so the next version has it. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] catalyst as non-root
On Wed, 2007-06-27 at 08:36 -0500, Andrew Gaffney wrote: Philipp Riegger wrote: Did anybody else try running catalyst as non-root and can give me hints on how to do it? We don't support catalyst running as non-root. If you want to do it, you're on your own. When catalyst was designed, nobody thought about running it as a non-root user. Well, I'm interested in adding support for running as a non-root user. The problem we're definitely going to face is device node creation. I've spoken with solar about using fakeroot and it is something that we would like to pursue for the next catalyst incarnation. Of course, we want catalyst to use fakeroot/chroot based on catalyst.conf settings, so we can't just replace the current code. This is a planned feature for the catalyst 2.1 branch, but it hasn't been started yet, so it'll be some time before we get to this point. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Livecd License
On Fri, 2007-06-29 at 07:36 -0500, Andrew Gaffney wrote: Luca Casagrande wrote: Hi to all! My livecd, will be used during a postgresql meeting to show some application. Organizing team, has asked me for the media's license; i told him that each application has got a own license and the full work can be release in gpl3. Am I correct? I doubt it. We can't really give you that answer as we have no idea what packages you are using on the LiveCD. You need to do the legwork of figuring out which license each package uses and if said license can be transferred to GPLv3. You can release the work as a compilation under a license of your choosing, so long as all the components that make up your work are licensed under a similar and compatible license. As an example, if you used a package that was under, for example, the ut2003 license, then you could not license under any GPL-compatible license since you would be unable to provide the source code for ut2003/4. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Problem with gnap/catalyst/genkernel and USE-flags
On Wed, 2007-06-20 at 13:23 +0300, Philipp Riegger wrote: On 20.06.2007, at 01:44, Chris Gianelloni wrote: The best/most tested method for changing USE is boot/kernel/$x/use in your spec file. Again, what does the [boot/kernel:x] in the documentation stand for? Huh? You call it $x above, the documentation calls it [boot/kernel:x] (e.g. boot/kernel/[boot/kernel:x]/use). That's completely inconsequential. I could have called it: boot/kernel/pinkponies/use and it still means the same thing. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] LiveCD with vanilla-sources
On Thu, 2007-06-14 at 11:35 -0500, Andrew Gaffney wrote: Genkernel can compile whatever sources /usr/src/linux points to. It's just that we use gentoo-sources for the releases and the default kernel configs that genkernel has are for gentoo-sources. However, they'll probably Just Work(TM) with vanilla-sources. If you use vanilla-sources, you cannot use squashfs. We don't recommend using vanilla-sources simply because we *never* test it. It will likely work, but I guarantee gentoo-sources will work. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] [PATCH] Use Portage's make.conf parser, allow user-specified extra bind mount points, allow absolute profile paths
On Mon, 2007-06-11 at 09:38 -0500, Charles Duffy wrote: If 'yall want me to split patch this out into multiple pieces, I'd be glad to do so. Please do. The things you've added all seem fine and would likely all be included, but I much prefer a one patch per change approach so we can easily back them out, if necessary. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] genkernel kernel location and version
On Mon, 2007-04-23 at 08:20 -0500, Andrew Gaffney wrote: You may have to use an overlay (or is it root_overlay?) to get the desired result. However, someone may have a better suggestion. I've never used the stage4 target myself. Well, you could disable kerncache since kerncache's code path is doing all of the changes you don't want. You have two choices here. Leave the sources (no kerncache) or not (kerncache). -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] genkernel kernel location and version
On Tue, 2007-04-24 at 00:23 +0200, Ramon van Alteren wrote: Chris Gianelloni wrote: Well, you could disable kerncache since kerncache's code path is doing all of the changes you don't want. You have two choices here. Leave the sources (no kerncache) or not (kerncache). Thanx, I wasn't aware that kerncache had that side-effect. Yeah, it is unfortunate, but the only way we can get kerncache capabilities with the current way we're doing things. I'm sure the kerncache could be modified to work a bit better, but the cache really needs to be all-inclusive, which is the reason it's done the way it is now in the first place. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Is this a bug?
On Wed, 2007-04-04 at 14:07 +, Nelson Batalha wrote: Uhh... it's all python/bash. What you have installed *is* the source code. There's nothing compiled in catalyst. ;] Lol, ok I may be ignorant, but not that much :P I just don't have a editor that searches all the catalyst files simultaneously. I'm looking for a way to print all the variables emerge uses, in the unmerge step, to see why isn't it finding the packages. targets/support/unmerge.sh targets/support/chroot-functions.sh (specifically run_emerge) -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Is this a bug?
On Wed, 2007-04-04 at 19:26 +, Nelson Batalha wrote: ROOT=/ emerge -pt -C sys-kernel/gentoo-sources One possible problem... don't use -V when doing your livecd-stage2 and see what you come up with... -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Is this a bug?
On Tue, 2007-04-03 at 17:24 +, Nelson Batalha wrote: Nelson wrote: After losing much sanity I realized the livecd's catalyst now makes seem to think my /dev/sdaX devices with a folder /livecd is The livecd. On targets/livecd-stage2/livecd-stage2-controller.sh: # Here is where we poke in our identifier touch $1/livecd If I add -${clst_livecd_type}-$2 (where $2 is the variable for the version stamp), where can the boot process be changed to look for this file? It's genkernel that looks for it, so you'd need to change genkernel. That being said, such an identifier can not have anything variable in it if you expect genkernel to find it. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Is this a bug?
On Tue, 2007-04-03 at 19:11 +, Nelson Batalha wrote: But since you have a boot option called nox (no X), and the boot options are specified by catalyst, perhaps it's possible for catalyst to write a boot option on the build, saying the livecd signature/version and genkernel would read it and look properly. Lol, it's not important, but an idea for someone to pick later? Uhh... I'd rather not touch anything of the sort. Passing things to genkernel from catalyst isn't as easy as you think it is and the best solution is likely the one we already have in place. Btw, I'm writting new docs before I forget http://gentoo-wiki.com/HOWTO_build_a_LiveCD_with_Catalyst_%28alternative%29 Great. =] I do have to warn you that the Release Engineering staff does not endorse *any* documentation placed on a public wiki of any kind. I have been working on updating the official documentation, which is the only thing that will be guaranteed to be even remotely correct. The last time I looked at the wiki, it told people to do really stupid things, which is why I simply don't recommend using any documentation not written by people intimately familiar with working with catalyst. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Is this a bug?
On Tue, 2007-04-03 at 20:47 +, Nelson Batalha wrote: So you're the unmerge is broken? Maybe I explained poorly. The packages are emerged, like gnome which in turn forces emerge gnome games. I want to unmerge gnome games on the last step. And gentoo-sources, etc. I could just delete their files, but would have to delete icons, menus, etc. Well, livecd/unmerge works fine here. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Some easy sugestions to implement, I hope
On Fri, 2007-03-30 at 16:32 +, Nelson Batalha wrote: You're very much right. Only I didn't demand anything cause right now I don't the time to learn python and submit, I was just replying to the new emails. Never crossed my mind to demand anything from a oss project :P, I just wanted to know if it was okfeasable. Just for clarification, I never thought you were demanding anything. As for your idea, it is definitely feasible, but not something I would *ever* want to implement. As much as you say it would be useful, I see it as a waste of *my* time, because it will encourage people to fiddle with their caches, and those people will be coming here, asking for help/support. What this means is that so long as I not only develop, but also *support* catalyst, I won't add anything that provides Release Engineering with no additional useful functionality but only serves to reduce my own availability for working on actual bugs. As much as I can see the needs of the casual/one-time user, I'm not designing a system catered to them. I think catalyst is too high of a learning curve already for a casual/one-time user, so I'd rather teach them the right way to go about using the tool if they have to do that much learning already. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Is this a bug?
On Sun, 2007-04-01 at 10:09 -0500, Andrew Gaffney wrote: This is a known bug with the currently stable version of genkernel in the tree. For the release, we've got a newer version marked stable. You can either mark the newer version stable in your snapshot or remove --unionfs-dev from 'livecd/gk_mainargs' in your livecd-stage2 spec file. People using --unionfs-dev should be shot in the face at close range with a nail gun. It specifically states that it is broken and for developer-use only. It is known to be completely buggy and only works when using the exact versions used by Release Engineering. I'll likely be removing this functionality entirely from genkernel simply because I'm tired of the bug reports caused by it after I've stated it isn't ready for use by pretty much anyone but myself. ;] I should state that explicitly. If your name is not Chris Gianelloni and you do not have commit access to the genkernel repository, you should not be using --unionfs-dev for any reason. If you do, don't file bugs. The real Chris Gianelloni won't support you on them and will likely be pissed off by your blatant impersonation attempts. That should do it. :P -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Some easy sugestions to implement, I hope
On Thu, 2007-03-29 at 01:05 +, Nelson Batalha wrote: Hi, You're using Gentoo and haven't figured out how to chroot? :-O Please forgive my assumptions about this, but last time I checked it was still required to do an installation ;) Yes,I don't mean just chroot, but there's more to it to simulate catalyst, like mounting lots of directories and setting environment variables and so on (who knows what it does, I can't read python and I'm a linux noob). I have a script for it, but has some problems unmounting dirs, and is probably incomplete. I really, really, REALLY, *REALLY* don't want to implement anything like that. As I said, it's a PITA enough supporting catalyst in all the millions of crazy ways people come up with using it. Making it easy, and thereby encouraging people, to poke around in the caches is the absolute last thing I want to do. I'm thinking about my sanity here. It is *very* rare that you should ever need to do anything in the cache itself. For troubleshooting, the information is available via other means, such as logs and temporary files that aren't deleted on failure. Mind you, I love your app, but it's as easy as its harder point - portage emerging lots of files wo/errors seems almost impossible. I've written dozens ebuilds in the overlay, emerged pkg's on my system. Still having errors. This is normal right? It does take Release Engineering upwards of a month to stabilize any given snapshot to build a release. Maybe there's a reason for that. ;] Hint: it isn't catalyst's fault (most of the time) -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] CVS location for spec files
On Thu, 2007-03-29 at 16:45 -0400, John Eckhart wrote: As of yet the only two problems I've encountered that weren't user error (there were plenty of those) were due to files that I believe are newer than the snapshot version (that's why I wanted to check the snapshot). I appreciate all the help, I'm growing to really appreciate catalyst. Release snapshots are complete and total hacks. Our current (second) release snapshot was taken on March 5th. Since then, we've injected upwards of 50 different ebuilds into it to resolve various bugs. We cannot publish the release snapshot until the release, as we often have security fixes in the snapshot that are not public (yet) and we could potentially get into some trouble for this if they were to be publicized. Unfortunately, until the release, your best course of action is to ask. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Generic stage1 and glibc
On Fri, 2007-03-23 at 11:48 -0400, John Eckhart wrote: I'm trying to build a generic stage1 using the stage3-x86-2006.1.tar.bz2 as a seed and the 20070319 portage snapshot. I can get all the way up to the building of glibc (glibc-2.5 as per the portage snapshot), which fails with the following error: * i386 CHOSTs are no longer supported. * Chances are you don't actually want/need i386. * Please read http://www.gentoo.org/doc/en/change-chost.xml !!! ERROR: sys-libs/glibc-2.5 failed. I wonder how the release team is producing the stage1's for 2007.0. Are they masking glibc-2.5 or have they moved the sub-architecture for stage1-x86 to something newer like i586? Neither. We're using chost: i486-pc-linux-gnu -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] [rfc] laying groundwork for cbuild
On Sat, 2007-01-20 at 00:31 -0500, Mike Frysinger wrote: On Wednesday 17 January 2007 10:07, Chris Gianelloni wrote: It was a livedvd spec, so I'll just post the top. Allow me to state that this worked perfectly before applying your patch, which, unlike my usual care, I applied without bothering to test... ;] whiner :p try the attached patch please Same problem. any plans to implement any sort of test suite ? Well, I usually try building at least livecd-stage1/livecd-stage2 with the changes prior to adding them to SVN. In the case of a patch touching some other location only, like a stage4 patch, I usually run through a stage4 before commit. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] [rfc] laying groundwork for cbuild
On Sat, 2007-01-20 at 00:31 -0500, Mike Frysinger wrote: On Wednesday 17 January 2007 10:07, Chris Gianelloni wrote: It was a livedvd spec, so I'll just post the top. Allow me to state that this worked perfectly before applying your patch, which, unlike my usual care, I applied without bothering to test... ;] whiner :p try the attached patch please any plans to implement any sort of test suite ? Ignore that last message. Your patch fixed it (at least for amd64/x86). -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] catalyst fails to umount dev
On Thu, 2007-01-18 at 01:05 +0100, Isidore Ducasse wrote: So catalyst should make the tarball by itself, shouldn't it? Why does it skip it on my machine? Anybody got a clue? Nope. The livecd-stage1 target does *not* produce a tarball. It merely builds up the cache for livecd-stage2 to use to build the ISO. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] [rfc] laying groundwork for cbuild
On Sun, 2006-12-31 at 10:09 -0500, Mike Frysinger wrote: attached patch redoes all of the logic for handling of chost and personality builds and inserts cbuild in all the right places any qualms before i take it further ? None here. Is this safe to apply $soon or do we need to wait for the whole shebang? I'm going to be rolling up 2.0.2 in the next couple weeks (at most) and am trying to get as much going as I can. What I have in SVN currently has been tested and all appears to work properly, so I'm getting close to another release. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Still on mouse problem..
On Wed, 2006-08-30 at 15:33 +0200, Luca Casagrande wrote: Hi to all! I am still checking mkxf86config.sh difference between 0.9.7 and 0.9.3: This is the section at line 159 from 0.9.3 (mouse working fine): # PS/2 bug: Some keyboards are incorrectly used as mice in XFree. :-( PSMOUSE=$(ls -l1 /dev/mouse* 2/dev/null | awk '/input/{print $NF ; exit 0}') if [ -n $PSMOUSE ]; And this is the same section in the 0.9.7 (mouse not working): # PS/2 bug: Some keyboards are incorrectly used as mice in XFree. :-( PSMOUSE=$(ls -l1 /dev/mouse* 2/dev/null | awk '/psaux/{print $NF ; exit 0}') if [ -n $PSMOUSE ]; As you can notice, the differences is between input and psaux. Before opening a bug i'd like to check your opinion. Please open a bug. Also, this can be fixed by enabling the legacy psaux interface in your kernel while you wait. -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] fsscript what order?
On Wed, 2006-08-30 at 12:35 -0400, Caleb Cushing wrote: at what point of boot does the fsscript load on the livecd? Never. The fsscript is run before your image is ever built. (...and catalyst gives output when this happens...) -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] xorg starting wrong user permissions.
On Wed, 2006-08-30 at 16:22 -0400, Caleb Cushing wrote: the file in question is the .dmrc this is not in /etc/skel and I do not create it. but for some reason it and the home directory are created with root ownership. I tried putting it what was suggested into an fsscript. however that didn't work. would the only way to solve this problem would to be to create them in the overlay? or with the script before hand so that they are user writeable? No. Get more specific with your fsscript. If the file is /home/foo/.dmrc, then specify it. chown 1000:100 /home/foo/.dmrc -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Re: Using masked packages in Catalyst .spec file
On Fri, 2006-08-25 at 20:36 +0200, Johan Tibell wrote: Branched from a question to Chris regarding using masked packages when building a minimal CD. Hi, I'm trying to build my own version of the 2006.0 minimal CD with madwifi support (I realize that madwifi is unstable). I've added a madwifi to the use flag section and madwifi-driver to the package section. I can't run catalyst -f livecd-stage1-minimal-madwifi.spec though since madwifi-driver is masked (error attached below). How do I unmask a package for the build? You cannot. You need to unmask it in your snapshot, or use portage_overlay to create a portage overlay. I found a portage_confdir variable in one of the .spec file that I can have point at a package.keywords file. Can you confirm that this is a way to use masked packages? You *can* but it isn't really advised. The best way really is via the snapshot, with the overlay as the second best, and confdir a distant last. -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Assign the user password
On Tue, 2006-08-22 at 20:39 +0200, Luca Casagrande wrote: Alle 18:50, martedì 22 agosto 2006, Chris Gianelloni ha scritto: On Tue, 2006-08-22 at 17:01 +0200, Luca Casagrande wrote: Now that much of the problem are solved, I am wondering if there's a way for setting the password of the user described in livecd-stage2.spec. I used to chroot in it and use passwd, but i wonder if there's a way to do it with catalyst. There's no method on catalyst. The simple answer is to do it in your fsscript, along with any other adjustments you need to make. Thx for your answer Chris...I have read a lot doc about bash, but i don't understand how set the password of the user. Can you post me an example? I just use sed to insert the password hash manually. I don't know of a decent commandline tool for changing the password non-interactively. -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Assign the user password
On Tue, 2006-08-22 at 13:42 -0500, Andrew Gaffney wrote: Luca Casagrande wrote: Thx for your answer Chris...I have read a lot doc about bash, but i don't understand how set the password of the user. Can you post me an example? echo 'username:password' | chpasswd ...or that... ;] -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Some error on my first build..
On Wed, 2006-08-23 at 11:20 +0200, Luca Casagrande wrote: But the error it's still there... Do you have livecd/users set? -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Re: Catalyst 2.0 weirdness
On Wed, 2006-08-23 at 06:58 +0200, Bardur Arantsson wrote: I think the livecd_stage2_target.py should probably give the user a slightly more informative error message ;). Feel free to file a bug with a patch. Remember that catalyst isn't really targeted for end-users (how many times do you hear that about a project) but instead Gentoo developers. We release it mostly for transparency reasons and because of our social contract. That being said, we do work bug reports, especially ones with patches. The only thing resembling the file name above is the source_subpath setting, but I'm not sure why that would get turned into a .tar.bz2 file name by catalyst when catalyst itself clearly expects a directory name. The change to the name of a .tar.bz2 bit still puzzles me and I'm not sure it's desirable behavior. It is. It is all due to seedcache and how it works. It first searches for a directory, then for a tarball. That's not saying that there isn't a bug in there somewhere, but the design is to use a directory first when options=seedcache and a tarball only otherwise for all targets *except* livecd-stage2. -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Problem with IceWM
On Wed, 2006-08-23 at 12:15 +0200, Luca Casagrande wrote: If i switch consolle, clear Xorg and startx as user, IceWm works fine (except mouse :) ). Wait... is gdm starting automatically, or have you done something to make it start? -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Some error on my first build..
On Wed, 2006-08-23 at 14:31 +0200, Luca Casagrande wrote: Alle 14:25, mercoledì 23 agosto 2006, Chris Gianelloni ha scritto: On Wed, 2006-08-23 at 11:20 +0200, Luca Casagrande wrote: But the error it's still there... Do you have livecd/users set? livecd/users: grass3d Strange... $ sed -i s:##STARTX:source /etc/profile su - grass3d -c startx: ~/.bashrc $ I get no error... -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Still on mouse problem...
On Wed, 2006-08-23 at 11:31 -0400, Caleb Cushing wrote: how do you set VIDEO_CARDS using catalyst? I didn't see that in the spec files? Well, you can set it multiple ways. You can either make your own profile and re-snapshot, which is probably the best way to do it, or you can add it to your envscript. -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Re: Catalyst 2.0 weirdness
On Wed, 2006-08-23 at 21:04 +0200, Bardur Arantsson wrote: I can't be bothered to open up a bug report, but I've attached a trivial patch. ... I think the logic in there still looks suspect though: AFAICT there may be other scenarios than the one I accidentally tried where 'unpack' may remain True while unpack_cmd does not get set to anything meaningful. With my patch this will result in the new error message as well -- I'm not sure how appropriate that is. Anyway, the patch certainly certainly shouldn't cause any regressions. The new message only triggers when an undefined variable would be accessed anyway. I'll have to test it. It sure would be nice if there were a bug report for this so that I didn't lose this email in all of my pre-release hustle. Maybe I'll happen to remember this email when I get to doing catalyst testing again. Who knows... -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Some error on my first build..
On Tue, 2006-08-22 at 09:53 -0500, Andrew Gaffney wrote: Luca Casagrande wrote: Thanks for your answer! I used: livecd/type: generic-livecd Ok. The error from 'mv' is harmless. It just means you aren't using gnome on your CD. The sed error I don't know about. It probably means there's an invalid sed statement somewhere in catalyst, but I don't know where. Right, the mv error is fine. The sed error is an actual error that only affects non-release builds, which is why I didn't find it before. It's been fixed in catalyst and will show up in 2.0.1, when it is released. Currently, it causes auto-login to not work properly for X on non-release builds. Fixing the sed at the end of targets/support/livecdfs-update.sh in catalyst should solve the issue. @@ -382,7 +382,7 @@ if [ -e /etc/startx ] then sed -i \ - s/##STARTX/source /etc/profile su - ${first_user} -c startx/ \ + s:##STARTX:source /etc/profile su - ${first_user} -c startx: \ /root/.bashrc fi -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Assign the user password
On Tue, 2006-08-22 at 17:01 +0200, Luca Casagrande wrote: Now that much of the problem are solved, I am wondering if there's a way for setting the password of the user described in livecd-stage2.spec. I used to chroot in it and use passwd, but i wonder if there's a way to do it with catalyst. There's no method on catalyst. The simple answer is to do it in your fsscript, along with any other adjustments you need to make. -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Still on mouse problem...
On Tue, 2006-08-22 at 17:21 +0200, Luca Casagrande wrote: Hi to all! I still can't have mouse works in X.. :( As i previous post, it's recognized by autoconfig and works in gpm but it remains in the middle of the screen inside GDM. I remembered that for 2006.0 livecd was used a symlink for mouse...I am wrong or this can help solve this problem? Modular X? Monolithic X? What USE flags? What INPUT_DEVICES? Which profile? We cannot help you without you providing the pertinent information. I can guarantee you that it works for me (and Andrew... and Danny...) on our 2006.1 stuff, so it is likely something different that you're doing. -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Some thoughts about NVIDIA and GPL issue...
On Fri, 2006-08-11 at 11:14 +0200, Luca Casagrande wrote: Hi to all! We all know that build nvidia or ati driver and then use it with our livecd is breaking the law.. My problem is that the livecd make use of 3d apps (Blender,Paraview,ecc..), so I am thinking about a solution: What about an init script runned before x-setup, that ask if you wanna use those driver? Well, this was already discussed on gentoo-dev mailing list some time ago, so it isn't a new idea. (Don't you think we would have already thought of this by now?) If yes, it will emerge nvidia or ati driver; if not it will pass by... The only issue with it is disk space. You will need an additional 100+ Megabytes of disk space to keep around your configured kernel sources. That is, unless someone takes the time to trace every single file in the kernel tree that gets touched by the drivers when building, and grabs just those files. Doing this would dramatically reduce the disk usage, but would dramatically increase the amount of work required to get a CD to build. -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] gensplash
On Wed, 2006-08-09 at 15:57 -0400, Caleb Cushing wrote: I've done a gensplash on desktop before.even got it to startup almost as fast as the official livecds by building it into the kernel. I've also gotten splash working in my livecd but it doesn't start till like 1/2 -3/4 through boot. how do I get it going faster? livecd/splash_type: gensplash livecd/splash_theme: livecd-2006.0 Honestly, no clue. This is something that we're experiencing on the 2006.1 CD builds, and we were unable to resolve it. Of course, we didn't spend too much time troubleshooting it, since other matters were more pressing, but I just wanted you to know that it isn't just you having the issue. -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] DHCP
On Wed, 2006-08-09 at 18:40 -0400, Caleb Cushing wrote: do I need to run dhcpcd to have the livecd get dhcp? it says probing for dhcp without it using auto configure however that doesn't seem to work? Is it detecting your network card? Have you read the Gentoo Handbook, specifically the portion on net-setup? The LiveCD environment doesn't always setup ethernet automatically. Sometimes, it simply fails. -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] kernel config and -p option
On Thu, 2006-08-10 at 14:33 +0200, Luca Casagrande wrote: Hi to all! I am on the livecd-stage2.spec step, but I got this error while building a wireless driver: ieee80211-1.1.13 requires the in-kernel version of the IEEE802.11 subsystem to be disabled (CONFIG_IEEE80211) On my first try of built this option was enabled on the kernel, so i removed it and now the config it's this: [..] # CONFIG_IEEE80211 is not set # CONFIG_IEEE80211_DEBUG is not set # CONFIG_IEEE80211_CRYPT_WEP is not set # CONFIG_IEEE80211_CRYPT_CCMP is not set # CONFIG_IEEE80211_CRYPT_TKIP is not set # CONFIG_IEEE80211_SOFTMAC is not set # CONFIG_IEEE80211_SOFTMAC_DEBUG is not set [..] On the next build with catalyst, I added the -p flag to clear any previous info, but got the same error. I am missing something or the --purge option doesn't remove previous kernel config? Thank you very much! Luca Did you also use -a? -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Catalyst and modular Xorg
On Tue, 2006-08-08 at 00:55 +0200, Luca Casagrande wrote: Hi to all! Thanks your great help i have successfully created a stage3 and now i am going to build livecd-stage1...Does i need to add all use flag for nvidia and ati to use proprietary driver, or it's enough to add xorg,nvidia-kernel,nvidia-glx and ati-drivers to package list without any related use flag? The binary drivers require a configured kernel, so they need to be built in livecd-stage2. Also, if you plan on giving this CD to *anyone* at all, you cannot use the binary drivers. It is against the law to distribute the binary drivers in a compiled form, since they violate the GPL and thereby copyright on the Linux kernel. -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Wiki document
On Tue, 2006-08-08 at 14:26 +0200, Luca Casagrande wrote: What do you think about? You're free to do what you want. However, after the 2006.1 release, I will be working on official documentation to be put up on the catalyst site. This will include an updated FAQ, catalyst 2.0 reference guide, and even a revival of the catalyst howto which explains how to do certain things. Except it to be something like the Gentoo Handbook for catalyst. The howto won't be a requirement for 2.0 going stable, but the reference guide will be. -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Wiki document
On Tue, 2006-08-08 at 12:01 -0400, Caleb Cushing wrote: snip update to the portage manpage. I'm not coming down on you, and gentoo generally has the best documentation around, however I notice many new tools (gentoo specific) are lacking it. Look. I really don't give a crap what else you've done. You are bitching on the CATALYST list about the CATALYST documentation. So I simply asked what you've contributed to the documentation. How about I give the answer, since I would have been the person you contributed it to. NOT A DAMN THING. Now, if you really feel like continuing this conversation, then feel free to contact me off-list. However, I'm not going to sit by and let some jackass bitch about the documentation for my project when he isn't contributing a damn thing to the effort. -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Re: Setting generation 1 and generation 2 Java VMs in stage 1
On Mon, 2006-08-07 at 11:34 +0200, Luca Casagrande wrote: Hi to all! Can you please tell me how you solved the problem..Each time i got a: [blocks B ] dev-java/java-config-wrapper (is blocking dev-java/java-config-1.2.11-r1) while creating the stage1.. Please, please, please, please, PLEASE quit calling livecd-stage1 stage1... They are not the same thing and it gets confusing otherwise. I have set all the package like the java upgrade guide in package.keywords... Umm... build your own set of seed stages. Honestly, you should have done this from the beginning, as it will reduce the number of problems that you'll have significantly. -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Catalyst Xorg Enabled livecd
On Thu, 2006-08-03 at 22:58 -0400, Caleb Cushing wrote: I'm new to catalyst first time using it. I've looked over the spec file templates and such. I get them but not the process... do I build a stage1 and then a stage2? and then 3 and so on? If you want the least amount of problems, yes. what packages are CRITICAL to be on the livecd? That completely depends on what you're doing. The big one is livecd-tools. However, it is so important, that we now force it on for livecd-stage1 builds. The only other one that I would say that you'll likely want/need will be coldplug/hwsetup. I want the packages that the cd runs to be current... like xorg 7, gcc 3.4.6-r1. Make your own snapshot. do I have to build my own tarball? instead of using the gentoo 2006.0 one for this? You should use a 2006.0 tarball only as a seed for your own stage1 tarball. When you use software that's 6 months old for your seed for livecd-stage1, you end up with way too many problems. -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] Catalyst Xorg Enabled livecd
On Fri, 2006-08-04 at 09:38 -0400, Caleb Cushing wrote: how can I set my cflags? I want -Os optimized instead of -O2 Well, use cflags in your stage1/stage2 spec file when building your own stage tarballs. I will warn you, that -Os doesn't always save space, and there are a few packages that break horribly when using -Os. -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] mkxf86config mouse detection
On Tue, 2006-08-01 at 14:25 -0700, Ted Kosan wrote: First, thanks again for the help I received on getting a Java VM on a LiveCD configured correctly :-) The next problem I am encountering involves mkxf86config ( version 0.9.7 ) not detecting the mouse when it generates the xorg.conf file. I am pretty sure I know the problem. Does anyone have some ideas on what I should try next? Yeah. I don't care what X -configure says. It works *so* totally differently that they aren't even comparable. You probably don't have hwsetup installed. Check the contents of your /etc/sysconfig directory (on the booted LiveCD) and see if it detects your mouse. If it doesn't neither will mkxf86config. Thanks in advance, No problem... glad I can help. =] -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] ANNOUNCE: catalyst 2.0
On Sat, 2006-07-29 at 18:57 -0700, M. Edward (Ed) Borasky wrote: Great news!! I haven't been to the Catalyst or Release Engineering sites recently ... is the documentation close to done? It hasn't even been started. It also won't be started until I return from Linux World in San Francisco, so at least 3 weeks. I don't consider the documentation necessary for the package to be unmasked, but it is a requirement for it to become stable and replace catalyst 1.x completely. The example spec files list all of the options available, as well as the comments in catalyst.conf, so you shouldn't be in the dark, or anything. -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
[gentoo-catalyst] ANNOUNCE: catalyst 2.0
As you can probably guess by the subject, catalyst 2.0 is now released. It is currently masked in the tree, and will be unmasked in ~2 days. The reason for the mask is so that I can get the message out. This completely replaces catalyst 1.x, and is no longer slotted differently. This also means that you will need to make a few changes in how you do things. For example, the default locations are now /etc/catalyst, /var/tmp/catalyst, and /usr/lib/catalyst, as they were with catalyst 1.x. You will also need to change your spec files to point to the new locations for your cdtar files, since they will have changed place. I'd like to thank everyone who has been involved in making this release. -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] How use livecd/fsscript and livecd/overlay
On Tue, 2006-07-25 at 17:52 +0200, Luca Casagrande wrote: Hi to all! I am moving to catalyst2 and for a simple livecd it's working very well. Now i need to do some advance operation: 1) I need to copy in the apache root the folder in /home/geko/pmapper; I need to use the livecd/overlay option, but i don't understand how set the place where put the folder (in my example /var/www/localhost/ht-docs). Should I use fsscript to copy the folder? Umm... you don't tell it where to put it... you make the overlay look the same... So you put your files in /home/geko/var/www/localhost/htdocs, instead... it's an overlay. Also, you don't want livecd/overlay, you want livecd/root_overlay. 2) How use fsscript to add option to kde menu? Just like you would do it from any shell script. A fsscript is *only* a shell script. It is copied and executed by catalyst near the end of your build. It can really be *anything* that's executable. For a KDE menu entry, I wouldn't waste my time with a fsscript and would instead build the .desktop file, then put it in livecd/root_overlay in the correct location, instead. -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: Re: [gentoo-catalyst] TEST REQUEST: catalyst-2.0_rc50
On Fri, 2006-07-21 at 19:42 +0200, Andreas Rückert wrote: Hi! Original-Nachricht Datum: Fri, 21 Jul 2006 12:39:55 -0400 Von: Chris Gianelloni [EMAIL PROTECTED] An: gentoo-catalyst@lists.gentoo.org Betreff: Re: [gentoo-catalyst] TEST REQUEST: catalyst-2.0_rc50 On Fri, 2006-07-21 at 10:56 +0200, Andreas Rückert wrote: I'd propose to add our results to http://gentoo-wiki.com/HOWTO_build_a_LiveCD I'm definitely *not* interested in adding anything to a wiki, where is can easily be screwed up by any passer-by. Since you always have the complete history, nothing is ever lost, no matter what any user does to the page.. Right but we also don't need to be pointing users to a page that at any point in time could say something like Andreas is a tool! in bright red 72 point font, either... ;] Having history is one thing, having control over the content is another. -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part
Re: [gentoo-catalyst] unionfs masked by x86 keyword?
On Thu, 2006-07-20 at 10:52 +, Alex wrote: Andreas Rückert wrote: Is it faster during the boot? I think it is, because it need not to copy the read-write image contents to tmpfs. I have never tested it, but that would make sense ;) Incorrect. Adding the unionfs package to the CD does exactly nothing but add the binaries and modules. It makes it only usable after the machine has booted. There is support for unionfs in genkernel, which means within the initramfs, but it is not functional right now and we simply don't have the time to look at it, since it's never been an advertised feature. Basically, if someone files a bug against it (without a patch) we mark it as INVALID and ask for a patch. We have contemplated many times simply removing the code altogether, since it is known to be broken, but we always hoped that we would scrape together some time to fix it. Using the in-genkernel code, it does boot faster. It also has a smaller memory footprint. -- Chris Gianelloni Release Engineering - Strategic Lead x86 Architecture Team Games - Developer Gentoo Linux signature.asc Description: This is a digitally signed message part