- Arch-general - lists.archlinux.org

[arch-general] A new approach to verify changes in the Linux kernel ABI
by Ponomarenko Andrey 14 Mar '16

14 Mar '16

Hello, I've released new versions of the abi-compliance-checker and abi-dumper tools that are now able to easily verify compatibility of changes in the Linux kernel ABI: https://github.com/lvc/ 1. The kernel should be compiled with CONFIG_DEBUG_INFO=y and CONFIG_DEBUG_INFO_REDUCED=n additional options. 2. Create ABI dump for vmlinux: abi-dumper vmlinux -vnum 4.4.5 -kernel-export -o ABI.dump (warning: may take about 5Gb of RAM and about half an hour to analyze debug info on a powerful computer) 3. Compare ABI dumps to produce report: abi-compliance-checker -l kernel -bin -old ABI-0.dump -new ABI-1.dump The new -kernel-export option of the abi-dumper is introduced to dump only public exported symbols of the kernel, i.e. declared in the ksymtab/ksymtab_gpl binary sections + system calls. I've applied this approach on the recent releases of the Linux kernel (2.6.36-4.4.5) as an example and prepared the compatibility table here: http://abi-laboratory.pro/tracker/timeline/linux/ The distribution specific changes in the kernel can be verified in the same way to prevent/minimize ABI differences with upstream or between sequential releases. Thank you.

1 0

[arch-general] Why does virtualbox-host-dkms have a hard dependency on linux-headers?
by Kyle Terrien 14 Mar '16

14 Mar '16

Hello, I upgraded my desktop about 10 minutes ago, and it looks like virtualbox-host-dkms gained a new dependency: linux-headers. On this desktop, I have only the linux-lts kernel and its accompanying linux-lts-headers installed, so why is linux-headers necessary? In the pacman -Syu output, it looks like dkms gives up when trying to build modules for the linux (4.4) kernel. However, it still builds the modules for linux-lts. Why does virtualbox-host-dkms have a hard dependency on linux-headers? --Kyle [2016-03-13 10:50] [ALPM] running '70-dkms-install.hook'... [2016-03-13 10:50] [ALPM-SCRIPTLET] ==> Kernel 4.4.5-1-ARCH modules are missing. Nothing will be done for this kernel! [2016-03-13 10:50] [ALPM-SCRIPTLET] ==> You have to install the matching kernel package to use dkms [2016-03-13 10:50] [ALPM-SCRIPTLET] [2016-03-13 10:50] [ALPM-SCRIPTLET] Creating symlink /var/lib/dkms/vboxhost/5.0.16_OSE/source -> [2016-03-13 10:50] [ALPM-SCRIPTLET] /usr/src/vboxhost-5.0.16_OSE [2016-03-13 10:50] [ALPM-SCRIPTLET] [2016-03-13 10:50] [ALPM-SCRIPTLET] DKMS: add completed. [2016-03-13 10:50] [ALPM-SCRIPTLET] [2016-03-13 10:50] [ALPM-SCRIPTLET] Kernel preparation unnecessary for this kernel. Skipping... [2016-03-13 10:50] [ALPM-SCRIPTLET] [2016-03-13 10:50] [ALPM-SCRIPTLET] Building module: [2016-03-13 10:50] [ALPM-SCRIPTLET] cleaning build area... [2016-03-13 10:51] [ALPM-SCRIPTLET] make KERNELRELEASE=4.1.19-1-lts -C /usr/lib/modules/4.1.19-1-lts/build M=/var/lib/dkms/vboxhost/5.0.16_OSE/build............... [2016-03-13 10:51] [ALPM-SCRIPTLET] cleaning build area... [2016-03-13 10:51] [ALPM-SCRIPTLET] Kernel cleanup unnecessary for this kernel. Skipping... [2016-03-13 10:51] [ALPM-SCRIPTLET] [2016-03-13 10:51] [ALPM-SCRIPTLET] DKMS: build completed. [2016-03-13 10:51] [ALPM-SCRIPTLET] [2016-03-13 10:51] [ALPM-SCRIPTLET] vboxdrv.ko: [2016-03-13 10:51] [ALPM-SCRIPTLET] Running module version sanity check. [2016-03-13 10:51] [ALPM-SCRIPTLET] [2016-03-13 10:51] [ALPM-SCRIPTLET] Good news! Module version 5.0.16_OSE for vboxdrv.ko [2016-03-13 10:51] [ALPM-SCRIPTLET] exactly matches what is already found in kernel 4.1.19-1-lts. [2016-03-13 10:51] [ALPM-SCRIPTLET] DKMS will not replace this module. [2016-03-13 10:51] [ALPM-SCRIPTLET] You may override by specifying --force. [2016-03-13 10:51] [ALPM-SCRIPTLET] [2016-03-13 10:51] [ALPM-SCRIPTLET] vboxnetflt.ko: [2016-03-13 10:51] [ALPM-SCRIPTLET] Running module version sanity check. [2016-03-13 10:51] [ALPM-SCRIPTLET] [2016-03-13 10:51] [ALPM-SCRIPTLET] Good news! Module version 5.0.16_OSE for vboxnetflt.ko [2016-03-13 10:51] [ALPM-SCRIPTLET] exactly matches what is already found in kernel 4.1.19-1-lts. [2016-03-13 10:51] [ALPM-SCRIPTLET] DKMS will not replace this module. [2016-03-13 10:51] [ALPM-SCRIPTLET] You may override by specifying --force. [2016-03-13 10:51] [ALPM-SCRIPTLET] [2016-03-13 10:51] [ALPM-SCRIPTLET] vboxnetadp.ko: [2016-03-13 10:51] [ALPM-SCRIPTLET] Running module version sanity check. [2016-03-13 10:51] [ALPM-SCRIPTLET] [2016-03-13 10:51] [ALPM-SCRIPTLET] Good news! Module version 5.0.16_OSE for vboxnetadp.ko [2016-03-13 10:51] [ALPM-SCRIPTLET] exactly matches what is already found in kernel 4.1.19-1-lts. [2016-03-13 10:51] [ALPM-SCRIPTLET] DKMS will not replace this module. [2016-03-13 10:51] [ALPM-SCRIPTLET] You may override by specifying --force. [2016-03-13 10:51] [ALPM-SCRIPTLET] [2016-03-13 10:51] [ALPM-SCRIPTLET] vboxpci.ko: [2016-03-13 10:51] [ALPM-SCRIPTLET] Running module version sanity check. [2016-03-13 10:51] [ALPM-SCRIPTLET] - Original module [2016-03-13 10:51] [ALPM-SCRIPTLET] - This kernel never originally had a module by this name [2016-03-13 10:51] [ALPM-SCRIPTLET] - Installation [2016-03-13 10:51] [ALPM-SCRIPTLET] - Installing to /usr/lib/modules/4.1.19-1-lts/kernel/misc/ [2016-03-13 10:51] [ALPM-SCRIPTLET] [2016-03-13 10:51] [ALPM-SCRIPTLET] depmod... [2016-03-13 10:51] [ALPM-SCRIPTLET] [2016-03-13 10:51] [ALPM-SCRIPTLET] DKMS: install completed. [2016-03-13 10:51] [ALPM-SCRIPTLET] ==> Kernel 4.4.5-1-ARCH modules are missing. Nothing will be done for this kernel! [2016-03-13 10:51] [ALPM-SCRIPTLET] ==> You have to install the matching kernel package to use dkms

4 6

[arch-general] Compositing broken on laptop, Compton and xcompmgr
by Harley W 14 Mar '16

14 Mar '16

I've been using Compton on my laptop for about a year without any issues. Recently, I noticed that transparency stopped working, and even more recently, shadows stopped working. If I try calling Compton from the command line I get an error saying "Failed to make GLX context current". I tried to use xcompmgr, but shadows and transparency still don't work, though I don't get any error messages. My Compton command is compton -b -z -c -r 32 My xcompmgr command is xcompmgr -r 32 -o 0.75 -l -16 -t -16 -a -c On Compton, I tried multiple backends, none were successful. I am testing this within bspwm, and got the same error on xmonad as well. However, compiz seems to be working perfectly in my xfce sessions. Recently, WiFi broke on my laptop as well, and now I have to run rfkill unblock WiFi when I start my computer. I'm thinking due to the recent onslaught of issues that this may be a hardware problem, but the laptop is only 3 years old. Any idea what's wrong? Thanks, Harley Wiltzer

4 5

[arch-general] mount shows hardcoded stripe=4 mount option for ext* /boot fs
by Bastien Traverse 13 Mar '16

13 Mar '16

I first stumbled upon this two years ago and opened a thread in the forum to ask about it [1]. I never got an answer but lived happy and forgetful about it, until some days ago another user PMed me to know if I had ever found out why mount consistently showed `stripe=4` as a mount option for the `/boot` partition, whether or not we leave it in /etc/fstab of use `defaults`. This piqued my interest again, so I'm submitting it here in hope somebody can enlighten us. Please see the original post for all the details. I can confirm that to this day, `$ mount | grep stripe` still returns /dev/mapper/boot-arch on / type ext4 (rw,stripe=4,data=ordered) but `# journalctl -b | grep stripe` returns kernel: EXT4-fs (dm-$): mounted filesystem with ordered data mode. Opts: (null) Cheers [1] https://bbs.archlinux.org/viewtopic.php?id=176092

3 3

[arch-general] stunnel
by flow 12 Mar '16

12 Mar '16

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 * stunnel.service - SSL tunnel for network daemons Loaded: loaded (/usr/lib/systemd/system/stunnel.service; enabled; vendor preset: disabled) Active: failed (Result: exit-code) Process: 583 ExecStart=/usr/bin/stunnel (code=exited, status=1/FAILURE) stunnel[583]: [!] /etc/stunnel/stunnel.conf:105: "[zxy]": Failed to initialize SSL systemd[1]: stunnel.service: Control process exited, code=exited status=1 systemd[1]: Failed to start SSL tunnel for network daemons. systemd[1]: stunnel.service: Unit entered failed state. systemd[1]: stunnel.service: Failed with result 'exit-code'. - -- Have a nice day! ~~~ This PGP signature only certifies the sender and date of the message. It implies no approval from the administrators of nym.mixmin.net. Date: Sat Mar 12 18:53:42 2016 GMT From: flow(a)nym.mixmin.net -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iEYEARECAAYFAlbkZbYACgkQViYZwngkfDurMQCgjD/uIQx9IRsWB5h7Z5atvpVO amAAnRMwpPg6MEuYLYUBAlbjXqsL8BC7 =hCCW -----END PGP SIGNATURE-----

3 2

[arch-general] anything in release notes cover this?
by Jude DaShiell 11 Mar '16

11 Mar '16

DKMS make.log for rt5592sta_linux_patched-dkms-2.6.0.0 for kernel 4.4.5-1-ARCH (x86_64) Thu Mar 10 23:37:22 EST 2016 make -C tools make[1]: Entering directory '/var/lib/dkms/rt5592sta_linux_patched-dkms/2.6.0.0/build/tools' gcc -g bin2h.c -o bin2h make[1]: Leaving directory '/var/lib/dkms/rt5592sta_linux_patched-dkms/2.6.0.0/build/tools' /var/lib/dkms/rt5592sta_linux_patched-dkms/2.6.0.0/build/tools/bin2h cp -f os/linux/Makefile.6 /var/lib/dkms/rt5592sta_linux_patched-dkms/2.6.0.0/build/os/linux/Makefile make -C /lib/modules/4.4.3-1-ARCH/build SUBDIRS=/var/lib/dkms/rt5592sta_linux_patched-dkms/2.6.0.0/build/os/linux modules make[1]: *** /lib/modules/4.4.3-1-ARCH/build: No such file or directory. Stop. Makefile:357: recipe for target 'LINUX' failed make: *** [LINUX] Error 2 --

2 1

Re: [arch-general] [arch-dev-public] DKMS modules for virtualbox
by Joakim Hernberg 09 Mar '16

09 Mar '16

We could stop splitting kernel headers out of the kernel package? As a user I think it would be ace if we had binary modules for those that want them but I have recently started using dkms modules again, and find it clearly superior for out of tree modules. (I maintain a couple of kernels myself, and for instance creating/updating and building nvidia packages is really somewhat of a pain). Personally I'd like to see the extramodules patching in the kernel buildscripts go away too ;) -- Joakim

3 2

[arch-general] IP of my coworking space blocked in Arch Linux Archive
by Damian Nowak 09 Mar '16

09 Mar '16

Well, where should I report that to get it unblocked? Thanks -Damian Nowak www.virtkick.com

2 1

[arch-general] most efficient way to get linux kernel statistics
by Andre Schmidt 08 Mar '16

08 Mar '16

hello archers, thought i ask here first, before i try some kernel mailing list. i'm writing (for fun) a tiny daemon that sends linux usage (cpu, mem, net, etc.) statistics as efficiently as possible to another machine, repeatedly, to get "live" data. at the moment i'm simply sending /proc files, but they sometimes have too much data. so was wondering if there is a more efficient way to get (only parts of) the data thats available in /proc? for example, /proc/meminfo has all this info: MemTotal: 8051660 kB MemFree: 6046020 kB MemAvailable: 6843080 kB Buffers: 109116 kB Cached: 708336 kB SwapCached: 0 kB Active: 1332416 kB Inactive: 455428 kB Active(anon): 971040 kB Inactive(anon): 22340 kB Active(file): 361376 kB Inactive(file): 433088 kB Unevictable: 16 kB Mlocked: 16 kB SwapTotal: 1048572 kB SwapFree: 1048572 kB Dirty: 16 kB Writeback: 0 kB AnonPages: 970412 kB Mapped: 323384 kB Shmem: 23008 kB Slab: 71760 kB SReclaimable: 44860 kB SUnreclaim: 26900 kB KernelStack: 6544 kB PageTables: 22924 kB NFS_Unstable: 0 kB Bounce: 0 kB WritebackTmp: 0 kB CommitLimit: 5074400 kB Committed_AS: 3487604 kB VmallocTotal: 34359738367 kB VmallocUsed: 0 kB VmallocChunk: 0 kB HardwareCorrupted: 0 kB AnonHugePages: 413696 kB HugePages_Total: 0 HugePages_Free: 0 HugePages_Rsvd: 0 HugePages_Surp: 0 Hugepagesize: 2048 kB DirectMap4k: 168600 kB DirectMap2M: 8091648 kB but i only need: MemTotal: 8051660 kB MemAvailable: 6843080 kB Cached: 708336 kB event better would be if kernel would only tell me: 8051660 6843080 708336 if /proc is the only way to get this info, i wonder if creating a kernel module for this would be more efficient, or even possible? Cheers Andre Schmidt

4 8

Re: [arch-general] Missing Dependency virtualbox-host-dkms
by Jack O'Connor 08 Mar '16

08 Mar '16

> I didn't force the deps to linux-headers because you could use the kernel > (linux, linux-lts, linux-zen, linux-grsec, etc) you prefer (or all together). > I will add a message with the next release of virtualbox. Would it be possible to depend on a "Provides" that all the different kernel header packages share? Similar to how nvidia-libgl, catalyst-libgl, and mesa-libgl all provide the "libgl" dependency. That would solve the presumably-very-common case where users initially installed virtualbox without any host headers at all, and are now landing in a broken state by default. (Or would that mean that the different header packages conflict where they didn't before?) - Jack

2 2