SA1100: fix building of h3100
Subsystems (2)
PATCH 1/1 SA1100: fix building of h3100
Both iPaqs h3100 and h3600 currently share the same source file - h3600.c But Makefile builds it only if CONFIG_SA1100_H3600 selected, so selecting just CONFIG_SA1100_H3100 results in "no machine record defined" message and aborted compilation.
Fix it.
---
Well, it left unnoticed by me for a while, just because I always build kernels with _both_ machines compiled in.
arch/arm/mach-sa1100/Makefile | 1 + 1 files changed, 1 insertions(+), 0 deletions(-)
arch/arm/mach-sa1100/Makefile +1 -0
@@ -25,6 +25,7 @@ led-$(CONFIG_SA1100_CERF) += leds-cerf.o obj-$(CONFIG_SA1100_COLLIE) += collie.o +obj-$(CONFIG_SA1100_H3100) += h3600.o obj-$(CONFIG_SA1100_H3600) += h3600.o obj-$(CONFIG_SA1100_HACKKIT) += hackkit.o
Signature · 1 line
1.6.3.4
PATCH 1/1 SA1100: fix building of h3100
Quoted 4 lines · Dmitry Artamonow
Both iPaqs h3100 and h3600 currently share the same source file -
h3600.cButMakefilebuilds it only ifCONFIG_SA1100_H3600selected, so selecting justCONFIG_SA1100_H3100results in "no machine record defined" message and aborted compilation.
Maybe this is what SA1100_H3XXX was meant for? Since that symbol's now gone, we might as well fix it this way.
PATCH 1/1 SA1100: fix building of h3100
Quoted 8 lines · Russell King - ARM Linux
Quoted 4 lines · Dmitry Artamonow
Both iPaqs h3100 and h3600 currently share the same source file -
h3600.cButMakefilebuilds it only ifCONFIG_SA1100_H3600selected, so selecting justCONFIG_SA1100_H3100results in "no machine record defined" message and aborted compilation.Maybe this is what
SA1100_H3XXXwas meant for? Since that symbol's now gone, we might as well fix it this way.
Yes, thanks. To be clear - this is a fix meant for 2.6.32.
I think h3600.c should be split in the future anyway, so there's really no need to revive SA1100_H3XXX.
Signature · 2 lines
Best regards, Dmitry "MAD" Artamonow
PATCH 1/1 SA1100: fix building of h3100
In reply to Dmitry ArtamonowQuoted 6 lines · On Mon, 5 Oct 2009 21:50:34 +0400 Dmitry Artamonow <[email protected]> wrote:
Both iPaqs h3100 and h3600 currently share the same source file -
h3600.cButMakefilebuilds it only ifCONFIG_SA1100_H3600selected, so selecting justCONFIG_SA1100_H3100results in "no machine record defined" message and aborted compilation.Fix it.
Acked
Quoted 29 lines
Signed-off-by Dmitry Artamonow <[email protected]>---
Well, it left unnoticed by me for a while, just because I always build kernels with _both_ machines compiled in.arch/arm/mach-sa1100/Makefile | 1 + 1 files changed, 1 insertions(+), 0 deletions(-)arch/arm/mach-sa1100/Makefile +1
-0@@ -25,6 +25,7 @@ led-$(CONFIG_SA1100_CERF) += leds-cerf.o obj-$(CONFIG_SA1100_COLLIE) += collie.o +obj-$(CONFIG_SA1100_H3100) += h3600.o obj-$(CONFIG_SA1100_H3600) += h3600.o obj-$(CONFIG_SA1100_HACKKIT) += hackkit.oSignature · 7 lines
1.6.3.4 _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel at lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Signature · 1 line
Kristoffer Ericson <[email protected]>
PATCH 1/1 SA1100: fix building of h3100
Quoted 11 lines · Kristoffer Ericson
Quoted 6 lines · On Mon, 5 Oct 2009 21:50:34 +0400 Dmitry Artamonow <[email protected]> wrote:
Both iPaqs h3100 and h3600 currently share the same source file -
h3600.cButMakefilebuilds it only ifCONFIG_SA1100_H3600selected, so selecting justCONFIG_SA1100_H3100results in "no machine record defined" message and aborted compilation.Fix it.
Acked
If there are people seriously looking at H3[16]00 stuff, it would be a good idea to move it over to use the proper GPIO support rather than continue to rely upon its own version.
Also maybe splitting it into h3100.c and h3600.c, with (if any) common bits in h3xxx.c might be an idea?
PATCH 1/1 SA1100: fix building of h3100
Quoted 9 lines · On Tue, 6 Oct 2009 12:28:31 +0100 Russell King - ARM Linux…
Quoted 8 lines · Kristoffer Ericson
Quoted 5 lines · On Mon, 5 Oct 2009 21:50:34 +0400 Dmitry Artamonow <[email protected]> wrote:
Both iPaqs h3100 and h3600 currently share the same source file -
h3600.cButMakefilebuilds it only ifCONFIG_SA1100_H3600selected, so selecting justCONFIG_SA1100_H3100results in "no machine record defined" message and aborted compilation.
Quoted 7 lines
Quoted 3 lines
Quoted 1 line
Fix it.
Acked
If there are people seriously looking at H3[16]00 stuff, it would be a good idea to move it over to use the proper GPIO support rather than continue to rely upon its own version.
Im not looking specificly at the IPAQ stuff, but it comes in contact with my jornada work. I agree that moving it into general GPIO support should be the priority.
Quoted 3 lines
Also maybe splitting it into
h3100.candh3600.c, with (if any) common bits inh3xxx.cmight be an idea?
Signature · 1 line
Kristoffer Ericson <[email protected]>
PATCH 1/1 SA1100: fix building of h3100
Kristoffer Ericson a ?crit :
Quoted 7 lines
Quoted 3 lines
If there are people seriously looking at H3[16]00 stuff, it would be a good idea to move it over to use the proper GPIO support rather than continue to rely upon its own version.
Im not looking specificly at the IPAQ stuff, but it comes in contact with my jornada work. I agree that moving it into general GPIO support should be the priority.
I do have an iPAQ 3630 and am also interested in moving it to general GPIO support. Maybe I can help there?
Amicalement,
Signature · 1 line
Albert.
PATCH 1/1 SA1100: fix building of h3100
In reply to Russell King - ARM LinuxQuoted 3 lines · Russell King - ARM Linux
If there are people seriously looking at H3[16]00 stuff, it would be a good idea to move it over to use the proper GPIO support rather than continue to rely upon its own version.
Yes, that's that I'm working on - the plan is to get rid of those horrible abstracted ipaq egpios, and use gpiolib functions instead (reusing htc-egpio driver for gpio expander)
I'll try to finish and clean up my patches soon, and submit them for review, so hopefully they could get into 2.6.33
Quoted 2 lines
Also maybe splitting it into
h3100.candh3600.c, with (if any) common bits inh3xxx.cmight be an idea?
Yes, as I said, that's the direction I'd like to go too. After conversion to gpiolib and real gpios there probably will left not much common at all.
Signature · 2 lines
Best regards, Dmitry "MAD" Artamonow