Re: [oe-commits] org.oe.dev udev-118: go back in time to udev-115 which seems to work slightly better on some Zaurus models

Re: [oe-commits] org.oe.dev udev-118: go back in time to udev-115 which seems to work slightly better on some Zaurus models

[Graeme Gregory]

On Sat, Jun 14, 2008 at 12:56:43AM +0200, mickeyl commit wrote:
> udev-118: go back in time to udev-115 which seems to work slightly better on some Zaurus models
> Zaurus models: need to force downgrade, I don't want to raise PE here.
> 
> # rename "packages/udev/udev-118"
> #     to "packages/udev/udev-115"
> # 
> # rename "packages/udev/udev_118.bb"
> #     to "packages/udev/udev_115.bb"
> # 

Im sorry I have to call for a dissaprove on this one, there is no way
in hell this is the correct way to deal with a small problem on a couple
of machines.

Graeme

Re: [oe-commits] org.oe.dev udev-118: go back in time to udev-115 which seems to work slightly better on some Zaurus models

[Koen Kooi]

Graeme Gregory wrote:
> On Sat, Jun 14, 2008 at 12:56:43AM +0200, mickeyl commit wrote:
>> udev-118: go back in time to udev-115 which seems to work slightly better on some Zaurus models
>> Zaurus models: need to force downgrade, I don't want to raise PE here.
>>
>> # rename "packages/udev/udev-118"
>> #     to "packages/udev/udev-115"
>> #
>> # rename "packages/udev/udev_118.bb"
>> #     to "packages/udev/udev_115.bb"
>> #
>
> Im sorry I have to call for a dissaprove on this one, there is no way
> in hell this is the correct way to deal with a small problem on a couple
> of machines.

Same here, that is absolutely the wrong way to 'fix' the problem.

Re: [oe-commits] org.oe.dev udev-118: go back in time to udev-115 which seems to work slightly better on some Zaurus models

[Andrea Adami]

>with a small problem on a couple
>of machines.

small problem = no touchscreen!

BTW you could even move to 119 (no evident problems).

Andrea

Re: [oe-commits] org.oe.dev udev-118: go back in time to udev-115 which seems to work slightly better on some Zaurus models

[Koen Kooi]

Andrea Adami wrote:
>> with a small problem on a couple
>> of machines.
>
> small problem = no touchscreen!
>
> BTW you could even move to 119 (no evident problems).

So why wasn't 119 *added* so people can test, instead of *removing* 118 
without any discussion?

Re: [oe-commits] org.oe.dev udev-118: go back in time to udev-115 which seems to work slightly better on some Zaurus models

[Andrea Adami]

Koen,

no evident problems == same behaviour as 118

could you keep an eye on the oebug 4118?

TIA

Andrea

Re: [oe-commits] org.oe.dev udev-118: go back in time to udev-115 which seems to work slightly better on some Zaurus models

[Graeme Gregory]

On Sat, Jun 14, 2008 at 03:34:34PM +0200, Andrea Adami wrote:
> Koen,
> 
> no evident problems == same behaviour as 118
> 
> could you keep an eye on the oebug 4118?
> 
Well I tried to build for tosa, but Im falling over on the fact the
kernel doesnt build and libtool seems to be busted :-(

G

Re: [oe-commits] org.oe.dev udev-118: go back in time to udev-115 which seems to work slightly better on some Zaurus models

[Koen Kooi]

Graeme Gregory wrote:
> On Sat, Jun 14, 2008 at 03:34:34PM +0200, Andrea Adami wrote:
>> Koen,
>>
>> no evident problems == same behaviour as 118
>>
>> could you keep an eye on the oebug 4118?
>>
> Well I tried to build for tosa, but Im falling over on the fact the
> kernel doesnt build and libtool seems to be busted :-(

I tracked this problem down to the fact that both libtool-cross and 
libtool install stuff into staging.

A possible solution would be to mimic gcc(-cross) so libtool-cross 
creates libltdl packages and stages things and libtool doesn't. This 
does mean that both libtool-cross and libtool need to be the same version.

To get around the current problem:

bitbake -b packages/libtool/libtool-cross_2.2.4.bb

That will tell packaged-staging to (re)stage the correct headers and m4 
files.

regards,

Koen

Re: [oe-commits] org.oe.dev udev-118: go back in time to udev-115 which seems to work slightly better on some Zaurus models

[Gordon]

[-- Attachment #1: Type: text/plain, Size: 314 bytes --]

On 06/14/2008 12:06 PM, Graeme Gregory wrote:
> Well I tried to build for tosa, but Im falling over on the fact the
> kernel doesnt build and libtool seems to be busted :-(

There are some kernel patch conflicts in the dev branch. You need the 
attached patch for linux-rp_2.6.24 on tosa.

Regards
Gordon

[-- Attachment #2: linux-rp_2.6.24-tosa.patch --]
[-- Type: text/x-diff, Size: 1647 bytes --]

#
# old_revision [ed4564d099fea7663820669914d24a6886bfc907]
#
# patch "packages/linux/linux-rp_2.6.24.bb"
#  from [4782fd14fc4ea11a863598f899afcf59024885eb]
#    to [f17b8575b063ee52c7d87d257a422569645b3a88]
#
============================================================
--- packages/linux/linux-rp_2.6.24.bb	4782fd14fc4ea11a863598f899afcf59024885eb
+++ packages/linux/linux-rp_2.6.24.bb	f17b8575b063ee52c7d87d257a422569645b3a88
@@ -48,9 +48,9 @@ SRC_URI = "${KERNELORG_MIRROR}pub/linux/
            file://mtd-module.patch;patch=1;status=external \
            file://wm8750-treble.patch;patch=1;status=external \
            file://spitz_h_rewrite.patch;patch=1;status=external \
-           file://pxa2xx_udc-clock.patch;patch=1 \
-           file://sharpsl-rc-r1.patch;patch=1 \
-           file://sharpsl-rc-r2.patch;patch=1 \
+#           file://pxa2xx_udc-clock.patch;patch=1 \
+#           file://sharpsl-rc-r1.patch;patch=1 \
+#           file://sharpsl-rc-r2.patch;patch=1 \
            file://squashfs3.3.patch;patch=1;status=external \
            ${RPSRC}/logo_oh-r1.patch.bz2;patch=1;status=unmergable \
            ${RPSRC}/pxa-linking-bug.patch;patch=1;status=unmergable \
@@ -65,7 +65,7 @@ SRC_URI = "${KERNELORG_MIRROR}pub/linux/
            file://connectplus-prevent-oops-HACK.patch;patch=1;status=hack \
            file://binutils-buildid-arm.patch;patch=1 \
            file://versatile-armv6.patch;patch=1 \
-	   file://htcuni.patch;patch=1 \
+#	   file://htcuni.patch;patch=1 \
            file://defconfig-c7x0 \
            file://defconfig-hx2000 \
            file://defconfig-collie \