[U-Boot-Users] [PATCH] fix system config overwrite @ MPC834x
Kim,
during 83xx setup the "System I/O configuration register high" gets overwritten with user defined value if CFG_SICRH is defined.
Regarding to the MPC834x manual (Table 5-28 reve.1) bits 28+29 of SICRH must keep their reset value regardless of configuration.
On my board (using RGMII) those bits are set after reset - yet it's unclear where they come from.
The patch keeps both bits on MPC834x.
Cheers, Andre
Signed-off-by: Andre Schwarz andre.schwarz@matrix-vision.de --
cpu/mpc83xx/cpu_init.c | 8 +++++++- 1 files changed, 7 insertions(+), 1 deletions(-)
diff --git a/cpu/mpc83xx/cpu_init.c b/cpu/mpc83xx/cpu_init.c index fba5b02..66cc1c2 100644 --- a/cpu/mpc83xx/cpu_init.c +++ b/cpu/mpc83xx/cpu_init.c @@ -59,6 +59,8 @@ static void config_qe_ioports(void) */ void cpu_init_f (volatile immap_t * im) { + u32 tmp_sicrh = 0; + /* Pointer is writable since we allocated a register for it */ gd = (gd_t *) (CFG_INIT_RAM_ADDR + CFG_GBL_DATA_OFFSET);
@@ -181,7 +183,11 @@ void cpu_init_f (volatile immap_t * im)
/* System General Purpose Register */ #ifdef CFG_SICRH - im->sysconf.sicrh = CFG_SICRH; +#ifdef CONFIG_MPC834X + /* regarding to MPC34x manual rev.1 bits 28..29 must be preserved */ + tmp_sicrh = im->sysconf.sicrh & 0x0000000C; +#endif + im->sysconf.sicrh = CFG_SICRH | tmp_sicrh; #endif #ifdef CFG_SICRL im->sysconf.sicrl = CFG_SICRL;
MATRIX VISION GmbH, Talstraße 16, DE-71570 Oppenweiler - Registergericht: Amtsgericht Stuttgart, HRB 271090 Geschäftsführer: Gerhard Thullner, Werner Armingeon, Uwe Furtner
On Thu, 17 Apr 2008 19:28:17 +0200 Andre Schwarz andre.schwarz@matrix-vision.de wrote:
Kim,
Hello Andre,
I can't apply this:
Applying fix system config overwrite @ MPC834x error: patch failed: cpu/mpc83xx/cpu_init.c:59 error: cpu/mpc83xx/cpu_init.c: patch does not apply Patch failed at 0001. When you have resolved this problem run "git-am --resolved". If you would prefer to skip this patch, instead run "git-am --skip".
during 83xx setup the "System I/O configuration register high" gets overwritten with user defined value if CFG_SICRH is defined.
Regarding to the MPC834x manual (Table 5-28 reve.1) bits 28+29 of SICRH must keep their reset value regardless of configuration.
On my board (using RGMII) those bits are set after reset - yet it's unclear where they come from.
The patch keeps both bits on MPC834x.
Cheers, Andre
Signed-off-by: Andre Schwarz andre.schwarz@matrix-vision.de
fyi, commit message text you don't want applied in the tree history (such as "Kim," and "Cheers, Andre") goes here, below the '---' line.
/* System General Purpose Register */#ifdef CFG_SICRH
- im->sysconf.sicrh = CFG_SICRH;
+#ifdef CONFIG_MPC834X
- /* regarding to MPC34x manual rev.1 bits 28..29 must be preserved */
- tmp_sicrh = im->sysconf.sicrh & 0x0000000C;
+#endif
- im->sysconf.sicrh = CFG_SICRH | tmp_sicrh;
#endif
also, can you extend the ifdef to include CONFIG_MPC8313 in addition to the MPC834X? That's the only other one that could use this fix.
Thanks,
Kim
Kim,
doing a git-pull gives "Already up-to-date." The patch is produced with "git-diff --patch-with-stat cpu/mpc83xx/cpu_init.c"
Am I doing anything wrong ?
Since all my patches have problems in getting applied there's obviously a problem on my side ...
regards, Andre
Kim Phillips schrieb:
On Thu, 17 Apr 2008 19:28:17 +0200 Andre Schwarz andre.schwarz@matrix-vision.de wrote:
Kim,
Hello Andre,
I can't apply this:
Applying fix system config overwrite @ MPC834x error: patch failed: cpu/mpc83xx/cpu_init.c:59 error: cpu/mpc83xx/cpu_init.c: patch does not apply Patch failed at 0001. When you have resolved this problem run "git-am --resolved". If you would prefer to skip this patch, instead run "git-am --skip".
during 83xx setup the "System I/O configuration register high" gets overwritten with user defined value if CFG_SICRH is defined.
Regarding to the MPC834x manual (Table 5-28 reve.1) bits 28+29 of SICRH must keep their reset value regardless of configuration.
On my board (using RGMII) those bits are set after reset - yet it's unclear where they come from.
The patch keeps both bits on MPC834x.
Cheers, Andre
Signed-off-by: Andre Schwarz andre.schwarz@matrix-vision.de
fyi, commit message text you don't want applied in the tree history (such as "Kim," and "Cheers, Andre") goes here, below the '---' line.
/* System General Purpose Register */#ifdef CFG_SICRH
- im->sysconf.sicrh = CFG_SICRH;
+#ifdef CONFIG_MPC834X
- /* regarding to MPC34x manual rev.1 bits 28..29 must be preserved */
- tmp_sicrh = im->sysconf.sicrh & 0x0000000C;
+#endif
- im->sysconf.sicrh = CFG_SICRH | tmp_sicrh;
#endif
also, can you extend the ifdef to include CONFIG_MPC8313 in addition to the MPC834X? That's the only other one that could use this fix.
Thanks,
Kim
MATRIX VISION GmbH, Talstraße 16, DE-71570 Oppenweiler - Registergericht: Amtsgericht Stuttgart, HRB 271090 Geschäftsführer: Gerhard Thullner, Werner Armingeon, Uwe Furtner
On Fri, 18 Apr 2008 10:48:51 +0200 Andre Schwarz andre.schwarz@matrix-vision.de wrote:
Kim,
doing a git-pull gives "Already up-to-date." The patch is produced with "git-diff --patch-with-stat cpu/mpc83xx/cpu_init.c"
using git-commit and git-format-patch is recommended, but that's probably not the problem here. Have you followed the instructions for Thunderbird in linux-2.6/Documentation/email-clients.txt? You can test by using git-am on patches you email yourself.
Kim
Thanks - I'll give it another try on monday.
Cheers, André
Kim Phillips wrote:
On Fri, 18 Apr 2008 10:48:51 +0200 Andre Schwarz andre.schwarz@matrix-vision.de wrote:
Kim,
doing a git-pull gives "Already up-to-date." The patch is produced with "git-diff --patch-with-stat cpu/mpc83xx/cpu_init.c"
using git-commit and git-format-patch is recommended, but that's probably not the problem here. Have you followed the instructions for Thunderbird in linux-2.6/Documentation/email-clients.txt? You can test by using git-am on patches you email yourself.
Kim
MATRIX VISION GmbH, Talstraße 16, DE-71570 Oppenweiler - Registergericht: Amtsgericht Stuttgart, HRB 271090 Geschäftsführer: Gerhard Thullner, Werner Armingeon, Uwe Furtner
Kim Phillips wrote:
On Fri, 18 Apr 2008 10:48:51 +0200 Andre Schwarz andre.schwarz@matrix-vision.de wrote:
Kim,
doing a git-pull gives "Already up-to-date." The patch is produced with "git-diff --patch-with-stat cpu/mpc83xx/cpu_init.c"
using git-commit and git-format-patch is recommended, but that's probably not the problem here. Have you followed the instructions for Thunderbird in linux-2.6/Documentation/email-clients.txt? You can test by using git-am on patches you email yourself.
Kim
Hi Andre,
Another thing I do when something I think should work doesn't is to make a fresh clone of the git repository, apply my new patch to it, and regenerate the patch from that. Sometimes I get confused as to the state of my working directory, what has been applied to it, what I've pulled from, etc. (I used to be really sharp, but then I discovered beer and the edges are a little duller now.)
FWIIW, I keep a local pristine copy of the denx.de master repository (and others I am interested in), doing a nightly pull in a cron job. This allows me to make a new working clone of what should be a clean repository without hammering denx.de. Worst case, I hammer denx.de occasionally. :-/
Best regards, gvb
participants (4)
-
Andre Schwarz -
André Schwarz
-
Jerry Van Baren -
Kim Phillips