From: rcorley <rc...@ae...> - 2011-10-10 16:24:30
|
I've found the need to modify one of the kernel drivers. Small changes to just one of the files in ../git/drivers/net. I want to know if there is a way to, after running the bitbake command to build the initial kernel, modify the desired .c file and then have that file compiled and the kernel rebuilt? I'll take any pointers/tips. I've been looking at numerous postings regarding patch files, building without bitbake, etc. and want to get some guidance before engaging on another lengthy trek down a path. Just in case there is an "easy" way to do this. -corley -- View this message in context: http://old.nabble.com/question%3A-how-do-I-rebuild-the-kernel-without--c-clean-and-another-bitbake--tp32625680p32625680.html Sent from the Gumstix mailing list archive at Nabble.com. |
From: Chris W. <whi...@gm...> - 2011-10-10 21:47:28
|
I believe you should be able to do "bitbake -c compile -f <recipe name>" to force it to re-compile your kernel. You'll have to get the uImage file out of your "work" directory" though. -chris On Mon, Oct 10, 2011 at 11:24 AM, rcorley <rc...@ae...> wrote: > > I've found the need to modify one of the kernel drivers. Small changes to > just one of the files in ../git/drivers/net. > > I want to know if there is a way to, after running the bitbake command to > build the initial kernel, modify the desired .c file and then have that file > compiled and the kernel rebuilt? > > I'll take any pointers/tips. I've been looking at numerous postings > regarding patch files, building without bitbake, etc. and want to get some > guidance before engaging on another lengthy trek down a path. Just in case > there is an "easy" way to do this. > > -corley > -- > View this message in context: http://old.nabble.com/question%3A-how-do-I-rebuild-the-kernel-without--c-clean-and-another-bitbake--tp32625680p32625680.html > Sent from the Gumstix mailing list archive at Nabble.com. > > > ------------------------------------------------------------------------------ > All the data continuously generated in your IT infrastructure contains a > definitive record of customers, application performance, security > threats, fraudulent activity and more. Splunk takes this data and makes > sense of it. Business sense. IT sense. Common sense. > http://p.sf.net/sfu/splunk-d2dcopy1 > _______________________________________________ > gumstix-users mailing list > gum...@li... > https://lists.sourceforge.net/lists/listinfo/gumstix-users > |
From: rcorley <rc...@ae...> - 2011-10-11 17:50:26
|
Update: The bitbake -c compile -f virtual/kernel command worked as I hoped. I followed it with a bitbake -c build -f virtual/kernel to build the image file Too bad I had to post this one to figure it out. I took a bit of time online to see if this information was out there on the web. No salient results from a search on "bitbake -c build". Seems like this little tidbit is important (it is to me, at least). Sorry for filling your inbox, but wanted to post that this worked. -corley rcorley wrote: > > I've found the need to modify one of the kernel drivers. Small changes to > just one of the files in ../git/drivers/net. > > I want to know if there is a way to, after running the bitbake command to > build the initial kernel, modify the desired .c file and then have that > file compiled and the kernel rebuilt? > > I'll take any pointers/tips. I've been looking at numerous postings > regarding patch files, building without bitbake, etc. and want to get some > guidance before engaging on another lengthy trek down a path. Just in > case there is an "easy" way to do this. > > -corley > -- View this message in context: http://old.nabble.com/question%3A-how-do-I-rebuild-the-kernel-without--c-clean-and-another-bitbake--tp32625680p32629475.html Sent from the Gumstix mailing list archive at Nabble.com. |
From: Don Q. de la M. <qu...@du...> - 2011-10-11 19:39:07
|
On Tue, Oct 11, 2011 at 10:50 AM, rcorley <rc...@ae...> wrote: > The bitbake -c compile -f virtual/kernel command worked as I hoped. > > I followed it with a bitbake -c build -f virtual/kernel to build the image > file > > Too bad I had to post this one to figure it out. I took a bit of time > online to see if this information was out there on the web. No salient > results from a search on "bitbake -c build". > > Seems like this little tidbit is important (it is to me, at least). Please do add it to the Wiki. It won't help you but it would help some other poor sot in the future: http://wiki.gumstix.org/index.php?title=Main_Page Don Quixote -- Don Quixote de la Mancha Dulcinea Technologies Corporation Software of Elegance and Beauty http://www.dulcineatech.com qu...@du... |
From: Dave H. <dhy...@gm...> - 2011-10-13 06:08:23
|
On Tue, Oct 11, 2011 at 12:38 PM, Don Quixote de la Mancha <qu...@du...> wrote: > On Tue, Oct 11, 2011 at 10:50 AM, rcorley <rc...@ae...> wrote: >> The bitbake -c compile -f virtual/kernel command worked as I hoped. >> >> I followed it with a bitbake -c build -f virtual/kernel to build the image >> file >> >> Too bad I had to post this one to figure it out. I took a bit of time >> online to see if this information was out there on the web. No salient >> results from a search on "bitbake -c build". >> >> Seems like this little tidbit is important (it is to me, at least). > > Please do add it to the Wiki. It won't help you but it would help > some other poor sot in the future: So I recently discovered the same thing. This page on the wiki: http://wiki.gumstix.org/index.php?title=Kernel_Reconfiguration has a link to this page, which is where I saw the appropriate bitbake instructions to use. http://blogs.elphel.com/2009/12/openembeddedangstrom-kernel-workflow/ So adding the information to that page seems like a reasonable place (to me). The Kernel Reconfiguration page mentions doing a bitbake -c clean virtual/kernel which is typically overkill, even for a modified kernel configuration (altough there are probably esoteric occasions where it is warranted). Having both options presented seems like a reasonable thing. -- Dave Hylands Shuswap, BC, Canada http://www.davehylands.com |
From: Chris W. <whi...@gm...> - 2011-10-13 18:16:24
|
If you do a clean, you will lose any modifications you made to the kernel. I think the way you should do kernel mods would be something like: 1. build the unmodified kernel the first time use the existing recipe 2. do a "quilt new <patch name>" to create your new patch 3. do a "quilt add <filename>" to each file you edit/change 4. make your changes. 5. do "bitbake -c compile -f <recipe name>" to compile your changes and test them 6. when you like your changes, do a "quilt refresh" to update your patch file 7. copy the patch file into your recipe folder, and update the recipe to add you patch during build 8. now you should be able to do a "bitbake -c clean -f <recipe>" and "bitbake <recipe> to test your new recipe and your patches. -chris On Thu, Oct 13, 2011 at 1:08 AM, Dave Hylands <dhy...@gm...> wrote: > On Tue, Oct 11, 2011 at 12:38 PM, Don Quixote de la Mancha > <qu...@du...> wrote: >> On Tue, Oct 11, 2011 at 10:50 AM, rcorley <rc...@ae...> wrote: >>> The bitbake -c compile -f virtual/kernel command worked as I hoped. >>> >>> I followed it with a bitbake -c build -f virtual/kernel to build the image >>> file >>> >>> Too bad I had to post this one to figure it out. I took a bit of time >>> online to see if this information was out there on the web. No salient >>> results from a search on "bitbake -c build". >>> >>> Seems like this little tidbit is important (it is to me, at least). >> >> Please do add it to the Wiki. It won't help you but it would help >> some other poor sot in the future: > > So I recently discovered the same thing. > > This page on the wiki: > http://wiki.gumstix.org/index.php?title=Kernel_Reconfiguration > > has a link to this page, which is where I saw the appropriate bitbake > instructions to use. > http://blogs.elphel.com/2009/12/openembeddedangstrom-kernel-workflow/ > > So adding the information to that page seems like a reasonable place (to me). > > The Kernel Reconfiguration page mentions doing a bitbake -c clean > virtual/kernel which is typically overkill, even for a modified kernel > configuration (altough there are probably esoteric occasions where it > is warranted). Having both options presented seems like a reasonable > thing. > > -- > Dave Hylands > Shuswap, BC, Canada > http://www.davehylands.com > > ------------------------------------------------------------------------------ > All the data continuously generated in your IT infrastructure contains a > definitive record of customers, application performance, security > threats, fraudulent activity and more. Splunk takes this data and makes > sense of it. Business sense. IT sense. Common sense. > http://p.sf.net/sfu/splunk-d2d-oct > _______________________________________________ > gumstix-users mailing list > gum...@li... > https://lists.sourceforge.net/lists/listinfo/gumstix-users > |