| Message ID | 1356220309-22847-1-git-send-email-stefan.froberg@petroprogram.com |
|---|---|
| State | Rejected |
| Headers | show |
>>>>> "Stefan" == Stefan Fröberg <stefan.froberg@petroprogram.com> writes:
Stefan> Signed-off-by: Stefan Fröberg <stefan.froberg@petroprogram.com>
Stefan> ---
Stefan> package/Config.in | 1 +
Stefan> package/libpciaccess/Config.in | 6 ++++++
Stefan> package/libpciaccess/libpciaccess.mk | 12 ++++++++++++
Stefan> 3 files changed, 19 insertions(+), 0 deletions(-)
Stefan> create mode 100644 package/libpciaccess/Config.in
Stefan> create mode 100644 package/libpciaccess/libpciaccess.mk
Ehh, we already have package/x11r7/xlib_libpciaccess - Any reason you
cannot use that?
I don't think there's any issue with bumping the version to 0.13.1
23.12.2012 11:15, Peter Korsgaard kirjoitti: >>>>>> "Stefan" == Stefan Fröberg <stefan.froberg@petroprogram.com> writes: > Stefan> Signed-off-by: Stefan Fröberg <stefan.froberg@petroprogram.com> > Stefan> --- > Stefan> package/Config.in | 1 + > Stefan> package/libpciaccess/Config.in | 6 ++++++ > Stefan> package/libpciaccess/libpciaccess.mk | 12 ++++++++++++ > Stefan> 3 files changed, 19 insertions(+), 0 deletions(-) > Stefan> create mode 100644 package/libpciaccess/Config.in > Stefan> create mode 100644 package/libpciaccess/libpciaccess.mk > > Ehh, we already have package/x11r7/xlib_libpciaccess - Any reason you > cannot use that? Oh, sorry. I was doing xserver 1.10 for buildroot 2012.08 and it was complaining about missing libpciaccess at some point (or maybe it was one of it's dependency package that was complaining, can't remember anymore) Didn't notice that there was already libpciaccess. Regards Stefan > I don't think there's any issue with bumping the version to 0.13.1 >
>>>>> "Stefan" == Stefan Fröberg <stefan.froberg@petroprogram.com> writes:
Hi,
Stefan> I was doing xserver 1.10 for buildroot 2012.08 and it was
Stefan> complaining about missing
Getting Xorg bumped to 1.10 (or rather all X packages to x11r7.7) would
be very good:
http://www.x.org/releases/X11R7.7/changelog.html
Please send (invidual) patches.
Hi Peter 23.12.2012 15:58, Peter Korsgaard kirjoitti: >>>>>> "Stefan" == Stefan Fröberg <stefan.froberg@petroprogram.com> writes: > Hi, > > Stefan> I was doing xserver 1.10 for buildroot 2012.08 and it was > Stefan> complaining about missing > > Getting Xorg bumped to 1.10 (or rather all X packages to x11r7.7) would > be very good: > > http://www.x.org/releases/X11R7.7/changelog.html > > Please send (invidual) patches. > Ok. I will. I sended already just few driver updates and libdrm bump and will continue after holidays. That mesa3d updating beyond 7.10.3 (that I have here) seems especially interesting challenge. Merry Christmas to everyone! Regards Stefan
Dear Stefan Fröberg, On Sun, 23 Dec 2012 16:04:16 +0200, Stefan Fröberg wrote: > That mesa3d updating beyond 7.10.3 (that I have here) seems especially > interesting challenge. Do you have more details about the issues involved in bumping mesa3d beyond 7.10.3 ? Thomas
Hi Thomas 23.12.2012 17:20, Thomas Petazzoni kirjoitti: > Dear Stefan Fröberg, > > On Sun, 23 Dec 2012 16:04:16 +0200, Stefan Fröberg wrote: > >> That mesa3d updating beyond 7.10.3 (that I have here) seems especially >> interesting challenge. > Do you have more details about the issues involved in bumping mesa3d > beyond 7.10.3 ? > > Thomas Well, if you try to bump it to 8.0.5 version, it will abort at some point. It borks at the time it tries to compile glsl. It tries to build helper program (builtin_compiler) with system python (shouldn't that be HOST_DIR/usr/bin/python ???) and then it crashes. I have included here the actual screenshot of the situation. Regards Stefan
27.12.2012 20:24, Stefan Fröberg kirjoitti: > Hi Thomas > > 23.12.2012 17:20, Thomas Petazzoni kirjoitti: >> Dear Stefan Fröberg, >> >> On Sun, 23 Dec 2012 16:04:16 +0200, Stefan Fröberg wrote: >> >>> That mesa3d updating beyond 7.10.3 (that I have here) seems especially >>> interesting challenge. >> Do you have more details about the issues involved in bumping mesa3d >> beyond 7.10.3 ? >> >> Thomas > Well, if you try to bump it to 8.0.5 version, it will abort at some point. > > It borks at the time it tries to compile glsl. > > It tries to build helper program (builtin_compiler) with system python > (shouldn't that be HOST_DIR/usr/bin/python ???) > and then it crashes. > > I have included here the actual screenshot of the situation. > > Regards > Stefan Update: The situation starts even earlier than 8.0.5. Right at version 7.11 And changing system python to buildroot python (MESA3D_CONF_ENV += PYTHON2= $(HOST_DIR)/usr/bin/python) didn't change anything. Regards Stefan
27.12.2012 21:08, Stefan Fröberg kirjoitti: > 27.12.2012 20:24, Stefan Fröberg kirjoitti: >> Hi Thomas >> >> 23.12.2012 17:20, Thomas Petazzoni kirjoitti: >>> Dear Stefan Fröberg, >>> >>> On Sun, 23 Dec 2012 16:04:16 +0200, Stefan Fröberg wrote: >>> >>>> That mesa3d updating beyond 7.10.3 (that I have here) seems especially >>>> interesting challenge. >>> Do you have more details about the issues involved in bumping mesa3d >>> beyond 7.10.3 ? >>> >>> Thomas >> Well, if you try to bump it to 8.0.5 version, it will abort at some point. >> >> It borks at the time it tries to compile glsl. >> >> It tries to build helper program (builtin_compiler) with system python >> (shouldn't that be HOST_DIR/usr/bin/python ???) >> and then it crashes. >> >> I have included here the actual screenshot of the situation. >> >> Regards >> Stefan > Update: > The situation starts even earlier than 8.0.5. Right at version 7.11 > And changing system python to buildroot python (MESA3D_CONF_ENV += > PYTHON2= $(HOST_DIR)/usr/bin/python) > didn't change anything. > Update2: Found out that this has nothing to do with python. The mesa glsl stuff building has changed since 7.10.3 so that it build some helper programs (like builtin_compiler and others) and it builds them wrongly with cross-compiler while it should use system compiler. Solution: Try to give correct variables (APP_CXX=g++ and APP_CC=gcc) to MESA3D_CONF_ENV and/or MESA3D_MAKE_ENV and see if problem goes away. If not, then make a patch that enforces the correct compiler for all those helper programs under src/glsl and src/glsl/glcpp
diff --git a/package/Config.in b/package/Config.in index 66445ff..d5e94db 100644 --- a/package/Config.in +++ b/package/Config.in @@ -431,6 +431,7 @@ source "package/libhid/Config.in" source "package/libiqrf/Config.in" source "package/libnfc/Config.in" source "package/libnfc-llcp/Config.in" +source "package/libpciaccess/Config.in" source "package/libusb/Config.in" source "package/libusb-compat/Config.in" source "package/libv4l/Config.in" diff --git a/package/libpciaccess/Config.in b/package/libpciaccess/Config.in new file mode 100644 index 0000000..ce5cbfc --- /dev/null +++ b/package/libpciaccess/Config.in @@ -0,0 +1,6 @@ +config BR2_PACKAGE_LIBPCIACCESS + bool "libpciaccess" + help + Library providing generic access to the PCI bus and devices + + http://cgit.freedesktop.org/xorg/lib/libpciaccess diff --git a/package/libpciaccess/libpciaccess.mk b/package/libpciaccess/libpciaccess.mk new file mode 100644 index 0000000..901c058 --- /dev/null +++ b/package/libpciaccess/libpciaccess.mk @@ -0,0 +1,12 @@ +############################################################# +# +# libpciaccess +# +############################################################# +LIBPCIACCESS_VERSION = 0.13.1 +LIBPCIACCESS_SOURCE = libpciaccess-$(LIBPCIACCESS_VERSION).tar.gz +LIBPCIACCESS_SITE = http://cgit.freedesktop.org/xorg/lib/libpciaccess/snapshot +LIBPCIACCESS_INSTALL_STAGING = YES + +$(eval $(autotools-package)) +
Signed-off-by: Stefan Fröberg <stefan.froberg@petroprogram.com> --- package/Config.in | 1 + package/libpciaccess/Config.in | 6 ++++++ package/libpciaccess/libpciaccess.mk | 12 ++++++++++++ 3 files changed, 19 insertions(+), 0 deletions(-) create mode 100644 package/libpciaccess/Config.in create mode 100644 package/libpciaccess/libpciaccess.mk