Re: [Angstrom-devel] [H5400-port] Merging .c files from arch/arm/mach-pxa/h5400 back to h5400.c
Hello, On Sun, 10 Feb 2008 22:43:32 +0100 "Michal Panczyk" <[EMAIL PROTECTED]> wrote: > Hello all > > A new year has come - so did the plans. Last year, before the > kernel-discuss went down (!!!), there was some work done for splitting > the baisic drivers of h5000 devices from the h5400.c file to separate > files. As the result some new files are now listed in > arch/arm/mach-pxa/h5400/ dir : > h5400_audio.c (removed by recent change in CVS) > h5400_bl.c > h5400_bt.c > h5400_lcd.c > h5400_wifi.c > > This year the plans have changed - the plan is to put everything back > to h5400.c file again... (see OE bug #3549 - just remember that the > temporary address change - bugs.openembedded.net). So I am going to > start work on bringing back the functions included in these files to > h5400.c. I got positive response from Paul Sokolovsky and Anton (who > spent some time on previous plan.). The work does not seem too > hard, as most of code is well documented (commented :P). I am going to > go by each file in alphabetical order and follow the style included in > h2200.c (as I worked with that a little and I know it is ok) and > submit patches in couple of next days. The plan is good. Note that it doesn't have to be done in one turn - I myself don't feel comfortable with merging all stuff into h4000.c, but anyway, it's just the matter of fact that we'll need to have that organization - one, at most 2 files per machine, to get stuff upstream. And indeed, one of the question here is to still have some consistent internal structure, I'm glad you're thinking about that, following h2200.c is a good plan. > > The hardest one for me seems to be the wifi driver - as I know that it > does not work in current state (I tried that form module). If I > have some problems with that one or any other one I will ask for some > help here or at kernel-bugs mailing list. > > I would like to get the patches tested even before going to OE.dev. I > know that most of people interested in looking through the patches > will know how to apply them and build proper image, but there are also > some people (who I guess have some more time) that would just like to > load the image and start testing. I think that information form the > test is quite valuable, so I would like to build and share an image > (probably console-image). I have started (again) the initial work on > some kind of test procedure available for everybody of testing support > of hardware components - > http://www.linuxtogo.org/gowiki/UniversalTestProcedure It is based on > my never sent report about current state of support of h5000 in > angstrom, but in the current state it is even unreadable (uh - how I > love formatting the wiki pages). In couple of days it is going to > look better, so I hope the future testers could use it quite > simple. Looks good. > I would be grateful for comment on that plan too - it seems to be > against the current 'policy' (every image has to go through > autobuilder), but it seems to be straight forward... Well, general availability images must go thru autobuilder. There's nothing wrong with building own testing images and sharing them with testers. Just please reserve share of your efforts towards making sure that the changes you made are pushed to the stable branch and then tested with autobuilder images. That does take effort, but is the only way to ensure sustainable progress, and not what we had before - multi-vector efforts of many people giving too little overall progress yet at the high price of doing the same stuff again and again. > > Also, Koen (I am not sure if You are subscribed to the list)- I have > seen Your comment at angstrom-devel > (http://article.gmane.org/gmane.linux.distributions.angstrom.devel/2058) > about getting Wifi to work. I don't want to mess up Your work, so if > You too have some comments pls share them here > > If there are any comments("go go go","the plan is wrong because", > "don't mess up with my code..." "leave this - there are move important > things to do in the port", "wait for 2.6.24 merger"), proposals please > submit them to h5400-port as most interested people are subscribed > there. > > Cheers > Michal Panczyk > ___ > H5400-port mailing list > [EMAIL PROTECTED] > https://www.handhelds.org/mailman/listinfo/h5400-port -- Best regards, Paul mailto:[EMAIL PROTECTED] ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
[Angstrom-devel] [RFC] Using IOSCHED_NOOP as the default for linux-handhelds
Hello, Looking for what more stuff can be made modular, I'd like to make IOSCHED_NOOP the default, builtin scheduler and make all other IO schedulers modules. Current default is the anticipatory one. From linux/block/Kconfig.iosched: config IOSCHED_NOOP bool default y ---help--- The no-op I/O scheduler is a minimal scheduler that does basic merging and sorting. Its main uses include non-disk based block devices like memory devices, and specialised software or hardware environments that do their own scheduling and require only minimal assistance from the kernel. config IOSCHED_AS tristate "Anticipatory I/O scheduler" default y ---help--- The anticipatory I/O scheduler is generally a good choice for most environments, but is quite large and complex when compared to the deadline I/O scheduler, it can also be slower in some cases especially some database loads. IMHO, noop's description is good enough for builtin flash and flash cards. I have no idea why IOSCHED_AS is used (pristine default is IOSCHED_CFQ), but it is the same as used by linux-rp, so I assume that's where it come from. So, I cc: Richard Purdie to ask if there're good grounds for that and to draw attention that it's a chance to recover some 5K for kernelspace-starving Z devices. -- Best regards, Paul mailto:[EMAIL PROTECTED] ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] [RFC] Adding explicit date to title of autobuilder RSS
Hello, On Wed, 6 Feb 2008 19:40:51 +0200 Paul Sokolovsky <[EMAIL PROTECTED]> wrote: > Hello, > > As stupid Firefox doesn't bother to render when dumping RSS, > I propose to add following patch to make it clear when this or that > image was built: > [] Committed and deployed: http://www.angstrom-distribution.org/unstable/autobuild/rss.php -- Best regards, Paul mailto:[EMAIL PROTECTED] ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] H3900
Hello, Please do not ignore suggestions to use mailing lists for communication. As I already mentioned that fact, I cc: the alist directly. On Thu, 7 Feb 2008 09:40:02 +0200 "Gints Polis" <[EMAIL PROTECTED]> wrote: > On Wed, Feb 6, 2008 at 10:50 PM, Paul Sokolovsky <[EMAIL PROTECTED]> > wrote: > > > Hello, > > > > On Wed, 6 Feb 2008 22:46:01 +0200 > > "Gints Polis" <[EMAIL PROTECTED]> wrote: > > > > > Hi, > > > > > > (I can use russian if you preffer) :) > > > > Nope, I prefer all project communication to happen thru public > > mailing lists ;-). > > > > > > > > I would like to test all possible options for H3900 device with > > > angstrom. So I would like to know is there a hope I can get > > > working my hardware with existing kernel source and what modules > > > I have to get compiled/check for: > > > > > > Compact Flash Card attached on ipaq sleeve > > > Wireless card attached on ipaq sleeve > > > sound > > > > All these require kernel work, so if you want to get it to work, > > that's where to start. > > > Ok. Do you have some link or somthing about actual state of these > device drivers in kernel source? http://handhelds.org/moin/moin.cgi/SupportedHandheldSummary is the authoritative location for status of any PocketPC device port, and informative location about other device ports. > > Gints > -- Best regards, Paul mailto:[EMAIL PROTECTED] ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] GPE panel dies
If it's a memory issue, swap would likely help. On Feb 10, 2008 10:24 AM, Paul Sokolovsky <[EMAIL PROTECTED]> wrote: > Hello, > > On Fri, 8 Feb 2008 02:11:19 -0500 > "SDG Systems" <[EMAIL PROTECTED]> wrote: > > > With Angstrom2007.12-stable, I have a problem with the GPE panel dying > > on a seemingly random basis. Has anyone else experienced this? I've > > tried both -r2 and current -r4, and tried two different platforms. The > > actual image that I'm building is 'x11-pimlico-image'. > > I've seen that too. In my list, that happens due to heavy memory > pressure X imposes even on 64Mb system. Just intuitively, no proofs, > sorry, except observation that there's always too little free memory > available and I saw it dying mostly on running something big. > > > Thanks. > > -- > > Todd > > [] > > -- > Best regards, > Paul mailto:[EMAIL PROTECTED] > > ___ > Angstrom-distro-devel mailing list > Angstrom-distro-devel@linuxtogo.org > http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel > ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] GPE panel dies
Hello, On Fri, 8 Feb 2008 02:11:19 -0500 "SDG Systems" <[EMAIL PROTECTED]> wrote: > With Angstrom2007.12-stable, I have a problem with the GPE panel dying > on a seemingly random basis. Has anyone else experienced this? I've > tried both -r2 and current -r4, and tried two different platforms. The > actual image that I'm building is 'x11-pimlico-image'. I've seen that too. In my list, that happens due to heavy memory pressure X imposes even on 64Mb system. Just intuitively, no proofs, sorry, except observation that there's always too little free memory available and I saw it dying mostly on running something big. > Thanks. > -- > Todd [] -- Best regards, Paul mailto:[EMAIL PROTECTED] ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] [RFC] stuff for r4 build
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Junqian Gordon Xu schreef: | To sum up: | | navit (need polish) | roadmap (need polish) | flite (correct mistake) | altboot | alsa-6000x-default.state.patch (not committed yet) Please send 5 e-mails to this list with [Review] in the subject, noting the revision numbers from .dev and the diff(s). You can find the revision numbers with mtn log and the easy way for the cumulative diff: mtn diff packages/ -r $(mtn head -b org.openembedded.angstrom-2007.12-stable | tail -n1 | awk '{print $1}') - -r $(mtn head -b org.openembedded.dev| tail -n1 | awk '{print $1}') regards, Koen - -- [EMAIL PROTECTED] will go go away in december 2007, please use [EMAIL PROTECTED] instead. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.5 (Darwin) iD8DBQFHrtZFMkyGM64RGpERAlu2AJ90A15uGDxdzoDSoUMNZemviqCX0QCfZ8uv Prfq6vyiL2yCTP2yOgaR9Ps= =EAdg -END PGP SIGNATURE- ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel