[arch-dev-public] [signoff] kernel26 2.6.27-1
I haven't been able to keep up many of my packages, and I'm sorry. At least this one was quick. kernel26 and the module packages are up in testing for both architectures. Here is the bad news: - gspcav1 has been merged into the kernel, the package can be dropped - aufs is broken again. I would like to drop it as it breaks constantly and we have to add three weird and old patches to the kernel so it works. We seem to be using only unionfs anyway. - catalyst is broken again. I am not motivated to fix it, so until someone else does it, it's broken - intel-536ep and intel-537 are broken. One error was easy to fix, but the other was not. I won't touch these packages and won't consider them blockers for the move to core. - ipw3945 is broken, I could probably fix it, but won't. I will remove ipw3945 from the repositories - I didn't look at openswan-klips, but I guess it still won't work. This has been broken since .24 or so, can we just drop it? - wlan-ng26 is broken. I could probably fix this as well, but I still doubt that anyone uses this driver now. If nobody fixes this in time for the core move, I'll drop the package. The rest either worked out of the box or I got it fixed. Along with the kernel and modules comes an updated v86d package that is a version bump plus a modification to the modprobe.d file that is necessary because an option of the uvesafb module has been renamed. The kernel has only been tested on x86_64 here, but it seems this is much smoother than the last few major kernel updates. Please give feedback.
Thomas Bächler schrieb:
I haven't been able to keep up many of my packages, and I'm sorry. At least this one was quick. kernel26 and the module packages are up in testing for both architectures.
I also added iwlwifi-5000-ucode to core, this will provide users with a Centrino 2 with wireless support out of the box on 2.6.27.
Thomas Bächler schrieb:
- catalyst is broken again. I am not motivated to fix it, so until someone else does it, it's broken
Okay, fixed it.
On Fri, 10 Oct 2008, Thomas Bächler wrote:
Thomas Bächler schrieb:
- catalyst is broken again. I am not motivated to fix it, so until someone else does it, it's broken
Okay, fixed it.
It looks like you forgot to commit the new catalyst patch in svn. -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean.
Eric Belanger escreveu:
On Fri, 10 Oct 2008, Thomas Bächler wrote:
Thomas Bächler schrieb:
- catalyst is broken again. I am not motivated to fix it, so until someone else does it, it's broken
Okay, fixed it.
It looks like you forgot to commit the new catalyst patch in svn.
Im having problems with the Load Modules step, but at least, that timer problem seems to be fixed (i was using 2.6.25, as .26 not worked at all for me). How can i debug it to see what is going on ? Tried to see dmesg, but it did not had anything strange. -- Douglas Soares de Andrade -- Python, Django, Zope e Plone == Archlinux Trusted User and Developer - dsa ** Quote: Old programmers never die; they exit to a higher shell.
Am Fri, 10 Oct 2008 20:06:28 +0200 schrieb Thomas Bächler <thomas@archlinux.org>:
Thomas Bächler schrieb:
- catalyst is broken again. I am not motivated to fix it, so until someone else does it, it's broken
Okay, fixed it.
Don't care about catalyst for the next few weeks/months/... AMD doesn't support our testing xorg-server that will soon move to extra. The development is so hidden that even RedHat and Ubuntu package maintainers don't get useful answers on the amd/ati beta list about planned releases and support for new xorg-server. I'm so disappionted that the closed beta mailing list isn't helpful in any way that I'm thinking about orphaning the package again and maybe even suggesting to push it back to AUR! Is there anybody who wants to adopt it? -Andy
On Fri, Oct 10, 2008 at 5:50 PM, Thomas Bächler <thomas@archlinux.org> wrote:
I haven't been able to keep up many of my packages, and I'm sorry. At least this one was quick. kernel26 and the module packages are up in testing for both architectures. Here is the bad news:
- gspcav1 has been merged into the kernel, the package can be dropped - aufs is broken again. I would like to drop it as it breaks constantly and we have to add three weird and old patches to the kernel so it works. We seem to be using only unionfs anyway. - catalyst is broken again. I am not motivated to fix it, so until someone else does it, it's broken - intel-536ep and intel-537 are broken. One error was easy to fix, but the other was not. I won't touch these packages and won't consider them blockers for the move to core. - ipw3945 is broken, I could probably fix it, but won't. I will remove ipw3945 from the repositories - I didn't look at openswan-klips, but I guess it still won't work. This has been broken since .24 or so, can we just drop it? - wlan-ng26 is broken. I could probably fix this as well, but I still doubt that anyone uses this driver now. If nobody fixes this in time for the core move, I'll drop the package.
The rest either worked out of the box or I got it fixed.
Along with the kernel and modules comes an updated v86d package that is a version bump plus a modification to the modprobe.d file that is necessary because an option of the uvesafb module has been renamed.
The kernel has only been tested on x86_64 here, but it seems this is much smoother than the last few major kernel updates. Please give feedback.
Looks fine here on 686. Tested iwl3945 wireless driver and uvesafb / v86d.
Am Freitag 10 Oktober 2008 17:50:36 schrieb Thomas Bächler:
Please give feedback.
This seems to be a nice kernel release. suspend is working again for me and sata_nv problems are gone, too. So I'll sign this off for i686 and x86_64. If those old intel modem drivers don't build anymore we should think about dropping support for them. I don't think they are widely used these days. -- Pierre Schmitz Clemens-August-Straße 76 53115 Bonn Telefon 0228 9716608 Mobil 0160 95269831 Jabber pierre@jabber.archlinux.de WWW http://www.archlinux.de
On Fri, 10 Oct 2008, Pierre Schmitz wrote:
Am Freitag 10 Oktober 2008 17:50:36 schrieb Thomas Bächler:
Please give feedback.
This seems to be a nice kernel release. suspend is working again for me and sata_nv problems are gone, too. So I'll sign this off for i686 and x86_64.
If those old intel modem drivers don't build anymore we should think about dropping support for them. I don't think they are widely used these days.
Boths system works fine. signoff for i686 and signoff (with nvidia) for x86_64. One issue: it seems that there is a new module for the pc speaker: snd_pcsp. When it is loaded, udev assign to it devices that are usually assigned to the sound card (/dev/dsp, etc). If it's loaded before the souncard module, /dev/dsp will be assign to the pc speaker and /dev/dsp1 will be assigned to the sound card causing some apps to fail using the sound card. the easiest way to fix it is to blacklist it and to reboot. That's how I fixed it. I don't have time to investigate that further. BTW, these's a related thread on the forum: http://bbs.archlinux.org/viewtopic.php?id=56724 -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean.
Eric Belanger schrieb:
One issue: it seems that there is a new module for the pc speaker: snd_pcsp. When it is loaded, udev assign to it devices that are usually assigned to the sound card (/dev/dsp, etc). If it's loaded before the souncard module, /dev/dsp will be assign to the pc speaker and /dev/dsp1 will be assigned to the sound card causing some apps to fail using the sound card. the easiest way to fix it is to blacklist it and to reboot. That's how I fixed it. I don't have time to investigate that further. BTW, these's a related thread on the forum: http://bbs.archlinux.org/viewtopic.php?id=56724
I added this driver. However only pcspkr is loaded here, snd_pcsp was never autoloaded. Both have a platform:pcspkr modalias. We can remove this module, but I thought it was rather cool to have.
Am Samstag 11 Oktober 2008 schrieb Thomas Bächler:
Eric Belanger schrieb:
One issue: it seems that there is a new module for the pc speaker: snd_pcsp. When it is loaded, udev assign to it devices that are usually assigned to the sound card (/dev/dsp, etc). If it's loaded before the souncard module, /dev/dsp will be assign to the pc speaker and /dev/dsp1 will be assigned to the sound card causing some apps to fail using the sound card. the easiest way to fix it is to blacklist it and to reboot. That's how I fixed it. I don't have time to investigate that further. BTW, these's a related thread on the forum: http://bbs.archlinux.org/viewtopic.php?id=56724
I added this driver. However only pcspkr is loaded here, snd_pcsp was never autoloaded. Both have a platform:pcspkr modalias. We can remove this module, but I thought it was rather cool to have. intel modem is fixed for both chipsets. rtl8187 seems to be not that stable as it was in .26 i'll observe it. The rest seems fine here.
greetings tpowa -- Tobias Powalowski Archlinux Developer & Package Maintainer (tpowa) http://www.archlinux.org tpowa@archlinux.org
HI also fixed wlan-ng26 compile greetings tpowa -- Tobias Powalowski Archlinux Developer & Package Maintainer (tpowa) http://www.archlinux.org tpowa@archlinux.org
Possible bug reports which could be inlcuded in next rebuild: http://bugs.archlinux.org/task/10900 http://bugs.archlinux.org/task/11201 greetings tpowa -- Tobias Powalowski Archlinux Developer & Package Maintainer (tpowa) http://www.archlinux.org tpowa@archlinux.org
2.6.27-2 will be going to testing in half an hour or so. Changes to -1 are: - Enabled watchdog timer drivers as modules (#11201) - Enable MSI/MSI-X support (#10900) I hope the last one doesn't force us to rebuild all modules again. Usually, you'll notice that nvidia or catalyst break. If there are any broken modules with -2 that worked with -1, we have to rebuild them all again, please tell me if such a problem occurs. If not, please sign off.
Thomas Bächler schrieb:
2.6.27-2 will be going to testing in half an hour or so. Changes to -1 are:
- Enabled watchdog timer drivers as modules (#11201) - Enable MSI/MSI-X support (#10900)
I hope the last one doesn't force us to rebuild all modules again. Usually, you'll notice that nvidia or catalyst break. If there are any broken modules with -2 that worked with -1, we have to rebuild them all again, please tell me if such a problem occurs. If not, please sign off.
Okay, apparently enabling MSI won't break external modules. However, they have to be rebuilt to be able to actually use MSI. IMO, this is not critical, but can be done for those modules that use it.
On Sun, Oct 12, 2008 at 9:09 AM, Thomas Bächler <thomas@archlinux.org> wrote:
Thomas Bächler schrieb:
2.6.27-2 will be going to testing in half an hour or so. Changes to -1 are:
- Enabled watchdog timer drivers as modules (#11201) - Enable MSI/MSI-X support (#10900)
I hope the last one doesn't force us to rebuild all modules again. Usually, you'll notice that nvidia or catalyst break. If there are any broken modules with -2 that worked with -1, we have to rebuild them all again, please tell me if such a problem occurs. If not, please sign off.
Okay, apparently enabling MSI won't break external modules. However, they have to be rebuilt to be able to actually use MSI. IMO, this is not critical, but can be done for those modules that use it.
i686 seems to be working fine here- a pretty generic setup however. -Dan
Am Sonntag 12 Oktober 2008 schrieb Dan McGee:
On Sun, Oct 12, 2008 at 9:09 AM, Thomas Bächler <thomas@archlinux.org> wrote:
Thomas Bächler schrieb:
2.6.27-2 will be going to testing in half an hour or so. Changes to -1 are:
- Enabled watchdog timer drivers as modules (#11201) - Enable MSI/MSI-X support (#10900)
I hope the last one doesn't force us to rebuild all modules again. Usually, you'll notice that nvidia or catalyst break. If there are any broken modules with -2 that worked with -1, we have to rebuild them all again, please tell me if such a problem occurs. If not, please sign off.
Okay, apparently enabling MSI won't break external modules. However, they have to be rebuilt to be able to actually use MSI. IMO, this is not critical, but can be done for those modules that use it.
i686 seems to be working fine here- a pretty generic setup however.
-Dan signoff x86_64 nvidia + raid1
-- Tobias Powalowski Archlinux Developer & Package Maintainer (tpowa) http://www.archlinux.org tpowa@archlinux.org
Am Sonntag 12 Oktober 2008 15:08:46 schrieb Thomas Bächler:
please sign off.
signed-off for both arches. Although I have no idea about the MSI stuff. :-) -- Pierre Schmitz Clemens-August-Straße 76 53115 Bonn Telefon 0228 9716608 Mobil 0160 95269831 Jabber pierre@jabber.archlinux.de WWW http://www.archlinux.de
On Mon, Oct 13, 2008 at 4:09 AM, Pierre Schmitz <pierre@archlinux.de> wrote:
Although I have no idea about the MSI stuff. :-)
cat /proc/interrupts You'll see something about MSI in there if your computer supports it, otherwise it will be business as usual with the PIC/APIC being used. -Dan
Am Montag 13 Oktober 2008 14:07:02 schrieb Dan McGee:
You'll see something about MSI in there if your computer supports it, otherwise it will be business as usual with the PIC/APIC being used.
Thats the problem. My computer should support it (pci 2.2 and pcie) but the output is still: cat /proc/interrupts CPU0 CPU1 0: 21 0 IO-APIC-edge timer 1: 8 4075 IO-APIC-edge i8042 7: 1 0 IO-APIC-edge 8: 0 0 IO-APIC-edge rtc0 9: 0 0 IO-APIC-fasteoi acpi 14: 238 109344 IO-APIC-edge pata_amd 15: 0 0 IO-APIC-edge pata_amd 16: 0 25 IO-APIC-fasteoi saa7146 (0) 17: 0 145 IO-APIC-fasteoi EMU10K1 18: 873 725649 IO-APIC-fasteoi nvidia 20: 190 76027 IO-APIC-fasteoi ohci_hcd:usb2 21: 0 2 IO-APIC-fasteoi ehci_hcd:usb1 22: 0 212 IO-APIC-fasteoi sata_nv 23: 1662 1171068 IO-APIC-fasteoi sata_nv, eth0 NMI: 0 0 Non-maskable interrupts LOC: 310691 332090 Local timer interrupts RES: 96205 31563 Rescheduling interrupts CAL: 6315 9220 function call interrupts TLB: 1268 748 TLB shootdowns TRM: 0 0 Thermal event interrupts THR: 0 0 Threshold APIC interrupts SPU: 0 0 Spurious interrupts ERR: 1 But there are some messages about msi in dmesg: dmesg | grep -i msi pci 0000:00:00.0: Enabling HT MSI Mapping pci 0000:00:0b.0: Enabling HT MSI Mapping pci 0000:00:0b.0: Found enabled HT MSI Mapping pci 0000:00:0c.0: Enabling HT MSI Mapping pci 0000:00:0c.0: Found enabled HT MSI Mapping pci 0000:00:0d.0: Enabling HT MSI Mapping pci 0000:00:0d.0: Found enabled HT MSI Mapping pci 0000:00:0e.0: Enabling HT MSI Mapping pci 0000:00:0e.0: Found enabled HT MSI Mapping pcieport-driver 0000:00:0b.0: found MSI capability pcieport-driver 0000:00:0c.0: found MSI capability pcieport-driver 0000:00:0d.0: found MSI capability pcieport-driver 0000:00:0e.0: found MSI capability -- Pierre Schmitz Clemens-August-Straße 76 53115 Bonn Telefon 0228 9716608 Mobil 0160 95269831 Jabber pierre@jabber.archlinux.de WWW http://www.archlinux.de
On Mon, 2008-10-13 at 14:39 +0200, Pierre Schmitz wrote:
Am Montag 13 Oktober 2008 14:07:02 schrieb Dan McGee:
You'll see something about MSI in there if your computer supports it, otherwise it will be business as usual with the PIC/APIC being used.
Thats the problem. My computer should support it (pci 2.2 and pcie) but the output is still:
I think your drivers should support MSI to be able to use it. My system has Intel ICH9R and Intel E1000 network, which use the ahci and e1000e drivers: [jan@jan ~]$ cat /proc/interrupts | grep -i MSI 761: 175705 175119 PCI-MSI-edge eth0 762: 193798 194868 PCI-MSI-edge ahci [jan@jan ~]$ dmesg | grep -i MSI pcieport-driver 0000:00:1c.0: found MSI capability pcieport-driver 0000:00:1c.1: found MSI capability pcieport-driver 0000:00:1c.2: found MSI capability pcieport-driver 0000:00:1c.3: found MSI capability pcieport-driver 0000:00:1c.4: found MSI capability
Jan de Groot schrieb:
On Mon, 2008-10-13 at 14:39 +0200, Pierre Schmitz wrote:
Am Montag 13 Oktober 2008 14:07:02 schrieb Dan McGee:
You'll see something about MSI in there if your computer supports it, otherwise it will be business as usual with the PIC/APIC being used. Thats the problem. My computer should support it (pci 2.2 and pcie) but the output is still:
I think your drivers should support MSI to be able to use it. My system has Intel ICH9R and Intel E1000 network, which use the ahci and e1000e drivers:
[jan@jan ~]$ cat /proc/interrupts | grep -i MSI 761: 175705 175119 PCI-MSI-edge eth0 762: 193798 194868 PCI-MSI-edge ahci
[jan@jan ~]$ dmesg | grep -i MSI pcieport-driver 0000:00:1c.0: found MSI capability pcieport-driver 0000:00:1c.1: found MSI capability pcieport-driver 0000:00:1c.2: found MSI capability pcieport-driver 0000:00:1c.3: found MSI capability pcieport-driver 0000:00:1c.4: found MSI capability
On my laptop I only see support in the iwl3945 driver. We could probably rebuild the nvidia driver so that it will support MSI as well, however that didn't help for Pierre.
Thomas Bächler schrieb:
2.6.27-2 will be going to testing in half an hour or so. Changes to -1 are:
- Enabled watchdog timer drivers as modules (#11201) - Enable MSI/MSI-X support (#10900)
I hope the last one doesn't force us to rebuild all modules again. Usually, you'll notice that nvidia or catalyst break. If there are any broken modules with -2 that worked with -1, we have to rebuild them all again, please tell me if such a problem occurs. If not, please sign off.
I'm moving this kernel and related modules to core and extra at midnight CEST (GMT+2), meaning in 6 hours. If anyone thinks there is a remaining problem that should be considered a blocker, please write to arch-dev-public or arch-general before that.
participants (9)
-
Andreas Radke
-
Dan McGee
-
Douglas Soares de Andrade
-
Eric Belanger
-
Jan de Groot
-
Pierre Schmitz
-
Thomas Bächler
-
Tobias Powalowski
-
Xavier