What's wrong with this code?
Hello,
Months ago I had a problem with network on Colibri PXA320. A applied all patches from Marek's web site http://openpxa.sourceforge.net/, but net didn't work at all in U-Boot and in Linux. When I tried to read/write into/from ethernet chip I/O registers, that was mapped in the memory, system hung.
See http://lists.infradead.org/pipermail/linux-arm-kernel/2010-February/009279.html
Solution was prompted by Daniel Mack (Thanks Daniel).
Add this code to lowlevel_init function in U-Boot:
//begin
ldr r0, =CKENA
ldr r1, [r0]
orr r1, r1, #(CKENA_4_NAND | CKENA_9_SMC) str r1, [r0]
ldr r1, [r0]
//end
After that network has worked in u-boot and in linux.
Also he said: "It did see this code, but it wasn't taken for upstream, for reasons I cannot remember..."
What's wrong with this code?
Thanks,
Best Regards
Dennis Semakin
Hi Dennis,
Quoted 18 lines
Hello,
Months ago I had a problem with network on Colibri PXA320. A applied all patches from Marek's web site http://openpxa.sourceforge.net/, but net didn't work at all in U-Boot and in Linux. When I tried to read/write into/from ethernet chip I/O registers, that was mapped in the memory, system hung.See http://lists.infradead.org/pipermail/linux-arm-kernel/2010-February/009279.html
Solution was prompted by Daniel Mack (Thanks Daniel).
Add this code to
lowlevel_initfunction in U-Boot:
//begin
ldr r0, =CKENA
ldr r1, [r0]
orr r1, r1, #(CKENA_4_NAND |CKENA_9_SMC) str r1, [r0]
ldr r1, [r0]
//end
Quoted 5 lines
After that network has worked in u-boot and in linux.
Also he said: "It did see this code, but it wasn't taken for upstream, for reasons I cannot remember..."What's wrong with this code?
Why don't you wrap this change into a patch and post it here[1] on the mailing list. If there are reasons not to merge it, the respective people will speak up.
Also note in the commit message that this fixes a real bug for you and what hardware you tested it on.
Cheers
Detlev
[1] http://www.denx.de/wiki/U-Boot/Patches
Signature · 6 lines
If we knew what it was we were doing, we wouldn't call it research.
-- Einstein
--
DENX Software Engineering GmbH, MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-40 Fax: (+49)-8142-66989-80 Email: dzu at denx.deHi
Quoted 4 lines
Why don't you wrap this change into a patch and post it here[1] on the mailing list. If there are reasons not to merge it, the respective people will speak up.
Because it's not my patch. I didn't write it. I did only copy and paste.
Best Regards,
Dennis
Hi Dennis,
Quoted 7 lines
Hi
Quoted 4 lines
Why don't you wrap this change into a patch and post it here[1] on the mailing list. If there are reasons not to merge it, the respective people will speak up.
Because it's not my patch. I didn't write it. I did only copy and paste.
That's not a big problem. Depending on how correct you want to be, you could either use "git commit --author <original author>" to commit tha change into your tree and then use git-send-email to post the patch here. The git toolchain will correctly keep the authors name even if you post the patches. Note that patches regularly get (re-)posted by people not being the original author. As long as it is correct from a copyright point of view, this is not a problem.
On such a small change however, you may also simply put a short "code originally by ..." in the commit message.
The _relevant_ part is that people here need a patch to do a proper review of the change you propose.
Cheers
Detlev
Signature · 5 lines
You live and learn -- DENX Software Engineering GmbH, MD: Wolfgang Denk & Detlev Zundel HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany Phone: (+49)-8142-66989-40 Fax: (+49)-8142-66989-80 Email: dzu at denx.de
Hi,
OK, thanks, I'll try
Best Regards,
Dennis Semakin