Debian kernel-config file for Linux 6.3
On Mon, 2023-05-08 at 07:48 +0200, Sedat Dilek wrote:
> Hi Ben and Debian kernel team,
>
> Today, I have seen on salsa.d.o there is some progress/update in
> master Git branch to use Linux 6.3
>
> Can you give help and/or instructions on how to generate a Debian
> kernel-config file?
> Alternatively point to a URL or provide such a kernel-config file?
You should be able to generate any .config file like this:
$ debian/rules debian/control-real
...
$ ARCH=$(dpkg --print-architecture) # or some other Debian
architecture
$ FEATURESET=none # or "rt" for builds with realtime
patches
$ FLAVOUR=... # whichever flavour you're interested
in
$ dpkg-architecture -a$ARCH -c make -f debian/rules.gen
setup_${ARCH}_${FEATURESET}_${FLAVOUR}
...
$ cat debian/build/build_${ARCH}_${FEATURESET}_${FLAVOUR}/.config
The setup rule is currently broken on master, so you'll need to apply
this patch first:
--- a/debian/rules.real
+++ b/debian/rules.real
@@ -345,7 +345,7 @@ binary_support: PACKAGE_ROOT =
/usr/share/$(PACKAGE_NAME)
dh_link $(PACKAGE_ROOT) /usr/src/$(PACKAGE_NAME)
+$(MAKE_SELF) install-base
-setup_image: $(STAMPS_DIR)/setup_$(ARCH)_$(FEATURESET)_$(FLAVOUR))
+setup_image: $(STAMPS_DIR)/setup_$(ARCH)_$(FEATURESET)_$(FLAVOUR)
build_image: $(STAMPS_DIR)/build_$(ARCH)_$(FEATURESET)_$(FLAVOUR)
--- END ---
Hi Ben,
this weekend I had some time to play with this.
You missed some PREREQS like you need a linux-x.y.z SOURCE...
What worked for me:
[ PREREQS + PREPS ]
[ /etc/apt/sources.list.d/debian-sources.list ]
# Debian/experimental sources
deb-src https://deb.debian.org/debian experimental main non-free contrib
deb-src https://deb.debian.org/debian rc-buggy main non-free contrib
- EOF-
apt-get update
cd $HOME/src/linux
apt-get source linux-source-6.3
cd linux-6.3.1/
git apply --verbose ../debian-rules_real.diff <--- Your DIFF (see also
patch below [1]).
[ GENERATE KERNEL-CONFIG ]
echo $ARCH $FEATURESET $FLAVOUR
amd64 none amd64
LC_ALL=C debian/rules debian/control-real
LC_ALL=C dpkg-architecture -a$ARCH -c make -f debian/rules.gen
setup_${ARCH}_${FEATURESET}_${FLAVOUR}
[ DIFFCONFIG ]
scripts/diffconfig /boot/config-6.3.0-0-amd64
debian/build/build_${ARCH}_${FEATURESET}_${FLAVOUR}/.config
+BUILD_SALT "6.3.0-0-amd64"
+MODULE_SIG_ALL n
+MODULE_SIG_KEY ""
+SYSTEM_TRUSTED_KEYS
"/home/dileks/src/linux/linux-6.3.1/debian/certs/debian-uefi-certs.pem"
Thanks.
Regards,
-Sedat-
[1]
https://salsa.debian.org/kernel-team/linux/-/commit/ad125a44dfa9fcfc4fcc79a697819a8259421484.patch
Re: Debian kernel-config file for Linux 6.3
On Mon, May 8, 2023 at 4:18 PM Ben Hutchings wrote:
>
> On Mon, 2023-05-08 at 07:48 +0200, Sedat Dilek wrote:
> > Hi Ben and Debian kernel team,
> >
> > Today, I have seen on salsa.d.o there is some progress/update in
> > master Git branch to use Linux 6.3
> >
> > Can you give help and/or instructions on how to generate a Debian
> > kernel-config file?
> > Alternatively point to a URL or provide such a kernel-config file?
[...]
> $ dpkg-architecture -a$ARCH -c make -f debian/rules.gen
> setup_${ARCH}_${FEATURESET}_${FLAVOUR}
> ...
> $ cat debian/build/build_${ARCH}_${FEATURESET}_${FLAVOUR}/.config
>
> The setup rule is currently broken on master, so you'll need to apply
> this patch first:
>
> --- a/debian/rules.real
> +++ b/debian/rules.real
> @@ -345,7 +345,7 @@ binary_support: PACKAGE_ROOT =
> /usr/share/$(PACKAGE_NAME)
> dh_link $(PACKAGE_ROOT) /usr/src/$(PACKAGE_NAME)
> +$(MAKE_SELF) install-base
>
> -setup_image: $(STAMPS_DIR)/setup_$(ARCH)_$(FEATURESET)_$(FLAVOUR))
> +setup_image: $(STAMPS_DIR)/setup_$(ARCH)_$(FEATURESET)_$(FLAVOUR)
>
> build_image: $(STAMPS_DIR)/build_$(ARCH)_$(FEATURESET)_$(FLAVOUR)
>
> --- END ---
>
Commit ad125a44 ("d/rules.real: Fix typo in setup_image target")
[1]
https://salsa.debian.org/kernel-team/linux/-/commit/ad125a44dfa9fcfc4fcc79a697819a8259421484
Re: Debian kernel-config file for Linux 6.3
On Mon, May 8, 2023 at 4:18 PM Ben Hutchings wrote:
>
> On Mon, 2023-05-08 at 07:48 +0200, Sedat Dilek wrote:
> > Hi Ben and Debian kernel team,
> >
> > Today, I have seen on salsa.d.o there is some progress/update in
> > master Git branch to use Linux 6.3
> >
> > Can you give help and/or instructions on how to generate a Debian
> > kernel-config file?
> > Alternatively point to a URL or provide such a kernel-config file?
>
> You should be able to generate any .config file like this:
>
> $ debian/rules debian/control-real
> ...
> $ ARCH=$(dpkg --print-architecture) # or some other Debian
> architecture
> $ FEATURESET=none # or "rt" for builds with realtime
> patches
> $ FLAVOUR=... # whichever flavour you're interested
> in
> $ dpkg-architecture -a$ARCH -c make -f debian/rules.gen
> setup_${ARCH}_${FEATURESET}_${FLAVOUR}
> ...
> $ cat debian/build/build_${ARCH}_${FEATURESET}_${FLAVOUR}/.config
>
> The setup rule is currently broken on master, so you'll need to apply
> this patch first:
>
> --- a/debian/rules.real
> +++ b/debian/rules.real
> @@ -345,7 +345,7 @@ binary_support: PACKAGE_ROOT =
> /usr/share/$(PACKAGE_NAME)
> dh_link $(PACKAGE_ROOT) /usr/src/$(PACKAGE_NAME)
> +$(MAKE_SELF) install-base
>
> -setup_image: $(STAMPS_DIR)/setup_$(ARCH)_$(FEATURESET)_$(FLAVOUR))
> +setup_image: $(STAMPS_DIR)/setup_$(ARCH)_$(FEATURESET)_$(FLAVOUR)
>
> build_image: $(STAMPS_DIR)/build_$(ARCH)_$(FEATURESET)_$(FLAVOUR)
>
> --- END ---
>
Hi Ben,
Thanks for the quick response and your help!
Much appreciated as I know you are in a stressed situation in
terminating Debian/bookworm release.
While digging through my own docs I found the attached document -
cannot say it's still up2date.
I will verify by generating a dot-config for a known Git tag like
debian/6.1.25-1 as you described.
Just see this morning linux-kernel v6.3.1 Debian packages in
Debian/rc-buggy (experimental):
root# apt-get install -t rc-buggy linux-image-6.3.0-0-amd64-unsigned
-o APT::Get::Upgrade-By-Source-Package=0 -y
( This time the installation went fine this way - with no (DKMS) errors. )
Have more fun.
Best regards,
-Sedat-
> Ben.
>
> --
> Ben Hutchings
> I haven't lost my mind; it's backed up on tape somewhere.
HELP-1: 4.2. Rebuilding official Debian kernel packages
HELP-1:
<https://kernel-handbook.alioth.debian.org/ch-common-tasks.html#s-common-official>
HELP-2: 4.2.5. Building packages for one flavour
HELP-2: <https://kernel-handbook.alioth.debian.org/ch-common-tasks.html#s4.2.5>
HELP-3: 4.5. Building a custom kernel from Debian kernel source
HELP-3:
<https://kernel-handbook.alioth.debian.org/ch-common-tasks.html#s-common-building>
[ PREPARE: DOWNLOAD DEBIAN-KERNEL SOURCES ]
mkdir -p ~/src/linux-kernel
cd ~/src/linux-kernel
apt-get source linux-image-$(uname -r)
[ PREPARE: ENTER BUILD-ENVIRONMENT ]
KVER="4.9.30"
cd linux-$KVER
[ PREPARE: GENERATE SETUP FOR ONE SINGLE_KERNEL_VARIANT ]
ARCH="amd64"
FEATURESET="none"
FLAVOUR="amd64"
SINGLE_KERNEL_VARIANT="${ARCH}_${FEATURESET}_${FLAVOUR}"
fakeroot make -f debian/rules.gen setup_$SINGLE_KERNEL_VARIANT
NOTE: This generates a debian/build/build_amd64_none_amd64/.config file.
[ EXAMPLE: DISABLE AND CHECK UNWANTED KCONFIG OPTIONS (OPTIONAL) ]
LINUX_CONFIG="debian/build/build_$SINGLE_KERNEL_VARIANT/.config"
scripts/config --file $LINUX_CONFIG --disable DEBUG_INFO
grep DEBUG_INFO $LINUX_CONFIG
NOTE: The usage of scripts/config is 'config options command' whereas --file is
an option and --disable a command (the order is important).
[ BUILD: GENERATE LINUX-IMAGE AND LINUX-HEADERS PACKAGES ]
MAX_JOBS=$(($(getconf _NPROCESSORS_ONLN)+1))
PARALLEL_MAKE_JOBS="$MAX_JOBS"
fakeroot make -j${PARALLEL_MAKE_JOBS} -f debian/rules.gen
binary-arch_$SINGLE_KERNEL_VARIANT
NOTE-1: This generates a linux-image-$(uname -r) and linux-headers-$(uname -r)
package.
NOTE-2: Using 'make -j' works in any case here to speedup the build.
NOTE-3: Using DEB_BUILD_OPTIONS=parallel=5 or DEBIAN_KERNEL_JOBS=5 was not
possible as both are defined in debian/rules.
[ BUILD: GENERATE MISSING LINUX-HEADERS-COMMON PACKAGE ]
fakeroot make -f debian/rules.gen binary-indep_${FEATURESET}
NOTE: binary-indep-featureset make-target in debian/rules.real invokes
install-common-headers_$(FEATURESET).
-sdi // 15-JUN-2017
Debian kernel-config file for Linux 6.3
Hi Ben and Debian kernel team, Today, I have seen on salsa.d.o there is some progress/update in master Git branch to use Linux 6.3 Can you give help and/or instructions on how to generate a Debian kernel-config file? Alternatively point to a URL or provide such a kernel-config file? Thanks in advance. Best regards, -Sedat- [1] https://salsa.debian.org/kernel-team/linux/-/tree/master
Re: [builddeb] dpkg: error processing package linux-image-6.2.0-rc4-1-amd64-clang15-kcfi (--configure)
This solved the issue:
# mv /etc/kernel/postinst.d/dkms /etc/kernel/postinst.d/dkms.orig
# LC_ALL=C dpkg -i
/home/dileks/src/linux/linux-image-6.2.0-rc4-1-amd64-clang15-kcfi_6.2.0~rc4-1~unstable+dileks1_amd64.deb
(Reading database ... 426963 files and directories currently installed.)
Preparing to unpack
.../linux-image-6.2.0-rc4-1-amd64-clang15-kcfi_6.2.0~rc4-1~unstable+dileks1_amd64.deb
...
Unpacking linux-image-6.2.0-rc4-1-amd64-clang15-kcfi
(6.2.0~rc4-1~unstable+dileks1) over (6.2.0~rc4-1~unstable+dileks1) ...
Setting up linux-image-6.2.0-rc4-1-amd64-clang15-kcfi
(6.2.0~rc4-1~unstable+dileks1) ...
update-initramfs: Generating /boot/initrd.img-6.2.0-rc4-1-amd64-clang15-kcfi
Generating grub configuration file ...
Found background image: /usr/share/images/desktop-base/desktop-grub.png
Found linux image: /boot/vmlinuz-6.2.0-rc4-1-amd64-clang15-kcfi
Found initrd image: /boot/initrd.img-6.2.0-rc4-1-amd64-clang15-kcfi
Found linux image: /boot/vmlinuz-6.1.0-1-amd64
Found initrd image: /boot/initrd.img-6.1.0-1-amd64
Warning: os-prober will be executed to detect other bootable partitions.
Its output will be used to detect bootable binaries on them and create
new boot entries.
Found Windows Recovery Environment on /dev/sdb3
done
# dpkg -l | grep linux-image | awk '/^(ii)/ {print $1 " " $2 " " $3}'
| column -t
ii linux-image-6.1.0-1-amd64 6.1.4-1
ii linux-image-6.2.0-rc4-1-amd64-clang15-kcfi 6.2.0~rc4-1~unstable+dileks1
ii linux-image-amd64
-Sedat-
Re: Upgrade of single Linux v6.1.0 (unsigned) package from Debian/experimental fails
Sorry, here the relevant part of ERROR quote: /etc/kernel/postinst.d/dkms: dkms: running auto installation service for kernel 6.1.0-0-amd64:Sign command: /lib/modules/6.1.0-0-amd64/build/scripts/sign-file Binary /lib/modules/6.1.0-0-amd64/build/scripts/sign-file not found, modules won't be signed For an unsigned linux-image you do NOT need the sign-file binary: Might be a DKMS issue? -Sedat-
Upgrade of single Linux v6.1.0 (unsigned) package from Debian/experimental fails
Hi, this was possible with linux-image-6.1.0-0-amd64-unsigned (6.1~rc8-1~exp1) but fails now with linux-image-6.1.0-0-amd64-unsigned (6.1.1-1~exp2): root# LC_ALL=C apt-get install -o APT::Get::Upgrade-By-Source-Package=0 -t experimental linux-image-6.1.0-0-amd64-unsigned -y ... Use 'apt autoremove' to remove them. Suggested packages: linux-doc-6.1 debian-kernel-handbook The following NEW packages will be installed: linux-image-6.1.0-0-amd64-unsigned 0 upgraded, 1 newly installed, 0 to remove and 306 not upgraded. Need to get 76.7 MB of archives. After this operation, 585 MB of additional disk space will be used. Get:1 https://incoming.debian.org/debian-buildd buildd-experimental/main amd64 linux-image-6.1.0-0-amd64-unsigned amd64 6.1.1-1~exp2 [76.7 MB] Fetched 76.7 MB in 1min 36s (797 kB/s) Selecting previously unselected package linux-image-6.1.0-0-amd64-unsigned. (Reading database ... 427330 files and directories currently installed.) Preparing to unpack .../linux-image-6.1.0-0-amd64-unsigned_6.1.1-1~exp2_amd64.deb ... Unpacking linux-image-6.1.0-0-amd64-unsigned (6.1.1-1~exp2) ... Setting up linux-image-6.1.0-0-amd64-unsigned (6.1.1-1~exp2) ... I: /vmlinuz.old is now a symlink to boot/vmlinuz-6.0.10-1-amd64-clang15-kcfi I: /initrd.img.old is now a symlink to boot/initrd.img-6.0.10-1-amd64-clang15-kcfi I: /vmlinuz is now a symlink to boot/vmlinuz-6.1.0-0-amd64 I: /initrd.img is now a symlink to boot/initrd.img-6.1.0-0-amd64 /etc/kernel/postinst.d/dkms: dkms: running auto installation service for kernel 6.1.0-0-amd64:Sign command: /lib/modules/6.1.0-0-amd64/build/scripts/sign-file Binary /lib/modules/6.1.0-0-amd64/build/scripts/sign-file not found, modules won't be signed Error! Your kernel headers for kernel 6.1.0-0-amd64 cannot be found at /lib/modules/6.1.0-0-amd64/build or /lib/modules/6.1.0-0-amd64/source. Please install the linux-headers-6.1.0-0-amd64 package or use the --kernelsourcedir option to tell DKMS where it's located. Error! One or more modules failed to install during autoinstall. Refer to previous errors for more information. failed! run-parts: /etc/kernel/postinst.d/dkms exited with return code 11 dpkg: error processing package linux-image-6.1.0-0-amd64-unsigned (--configure): installed linux-image-6.1.0-0-amd64-unsigned package post-installation script subprocess returned error exit status 1 Errors were encountered while processing: linux-image-6.1.0-0-amd64-unsigned E: Sub-process /usr/bin/dpkg returned an error code (1) This solves it: root@iniza:~/packages# LC_ALL=C apt-get install -o APT::Get::Upgrade-By-Source-Package=0 -t experimental linux-image-6.1.0-0-amd64-unsigned linux-headers-6.1.0-0-amd64 linux-headers-6.1.0-0-common -y ... Use 'apt autoremove' to remove them. The following NEW packages will be installed: linux-headers-6.1.0-0-amd64 linux-headers-6.1.0-0-common linux-kbuild-6.1 0 upgraded, 3 newly installed, 0 to remove and 306 not upgraded. 1 not fully installed or removed. Need to get 11.6 MB of archives. After this operation, 62.5 MB of additional disk space will be used. Get:1 https://incoming.debian.org/debian-buildd buildd-experimental/main amd64 linux-headers-6.1.0-0-common all 6.1.1-1~exp2 [9703 kB] Get:2 https://incoming.debian.org/debian-buildd buildd-experimental/main amd64 linux-kbuild-6.1 amd64 6.1.1-1~exp2 [816 kB] Get:3 https://incoming.debian.org/debian-buildd buildd-experimental/main amd64 linux-headers-6.1.0-0-amd64 amd64 6.1.1-1~exp2 [1086 kB] Fetched 11.6 MB in 18s (632 kB/s) Selecting previously unselected package linux-headers-6.1.0-0-common. (Reading database ... 432240 files and directories currently installed.) Preparing to unpack .../linux-headers-6.1.0-0-common_6.1.1-1~exp2_all.deb ... Unpacking linux-headers-6.1.0-0-common (6.1.1-1~exp2) ... Selecting previously unselected package linux-kbuild-6.1. Preparing to unpack .../linux-kbuild-6.1_6.1.1-1~exp2_amd64.deb ... Unpacking linux-kbuild-6.1 (6.1.1-1~exp2) ... Selecting previously unselected package linux-headers-6.1.0-0-amd64. Preparing to unpack .../linux-headers-6.1.0-0-amd64_6.1.1-1~exp2_amd64.deb ... Unpacking linux-headers-6.1.0-0-amd64 (6.1.1-1~exp2) ... Setting up linux-image-6.1.0-0-amd64-unsigned (6.1.1-1~exp2) ... /etc/kernel/postinst.d/dkms: dkms: running auto installation service for kernel 6.1.0-0-amd64:Sign command: /usr/lib/linux-kbuild-6.1/scripts/sign-file Signing key: /var/lib/dkms/mok.key Public certificate (MOK): /var/lib/dkms/mok.pub Building module: Cleaning build area... make -j4 KERNELRELEASE=6.1.0-0-amd64 -C /lib/modules/6.1.0-0-amd64/build M=/var/lib/dkms/virtualbox/7.0.4/build. Signing module /var/lib/dkms/virtualbox/7.0.4/build/vboxdrv/vboxdrv.ko Signing module /var/lib/dkms/virtualbox/7.0.4/build/vboxnetadp/vboxnetadp.ko Signing module /var/lib/dkms/virtualbox/7.0.4/build/vboxnetflt/vboxnetflt.ko Cleaning build area... vboxdrv.ko: Running module version sanity check. - Original module - No original module exists within this kernel -
Bug#1014319: depmod: WARNING: could not open modules.builtin.modinfo at /var/tmp/mkinitramfs_vBlw4a/lib/modules/5.18.0-2-amd64: No such file or directory
Hi, with the fix from Sven Joachim applied to mkinitramfs I do not see the warning of kmod-30. Feel free to add my: Tested-by: Sedat Dilek # Debian/unstable AMD64 Thanks. Regards, -Sedat-
Bug#982304: linux-headers-5.10.0-3-amd64: File include/generated/autoconf.h missed to compile virtualbox-dkms
If you like you can close this bug-report. linux-headers-5.10.0-4-amd64 version 5.10.19-1 ships the necessary files and I was able to build virtualbox-dkms kernel-modules: root# LC_ALL=C ll /lib/modules/5.10.0-4-amd64/updates/dkms/ total 852K drwxr-xr-x 2 root root 4.0K Mar 3 04:00 . drwxr-xr-x 3 root root 4.0K Mar 3 04:00 .. -rw-r--r-- 1 root root 755K Mar 3 04:00 vboxdrv.ko -rw-r--r-- 1 root root 21K Mar 3 04:00 vboxnetadp.ko -rw-r--r-- 1 root root 61K Mar 3 04:00 vboxnetflt.ko - Sedat -
Bug#982304: linux-headers-5.10.0-3-amd64: File include/generated/autoconf.h missed to compile virtualbox-dkms
I checked the filelist from [1]: # echo $files /usr/src/linux-headers-5.10.0-3-amd64/include/generated/asm-offsets.h /usr/src/linux-headers-5.10.0-3-amd64/include/generated/autoconf.h /usr/src/linux-headers-5.10.0- 3-amd64/include/generated/bounds.h /usr/src/linux-headers-5.10.0-3-amd64/include/generated/compile.h /usr/src/linux-headers-5.10.0-3-amd64/include/generated/package.h /usr/src/linux-headers-5.10.0-3-amd64/include/generated/timeconst.h /usr/src/linux-headers-5.10.0-3-amd64/include/generated/uapi/linux/version.h /usr/src/linux-headers -5.10.0-3-amd64/include/generated/utsrelease.h # LC_ALL=C ll $files ls: cannot access '/usr/src/linux-headers-5.10.0-3-amd64/include/generated/autoconf.h': No such file or directory -rw-r--r-- 1 root root 4.7K Feb 6 09:23 /usr/src/linux-headers-5.10.0-3-amd64/include/generated/asm-offsets.h -rw-r--r-- 1 root root 318 Feb 6 09:23 /usr/src/linux-headers-5.10.0-3-amd64/include/generated/bounds.h -rw-r--r-- 1 root root 337 Feb 6 09:23 /usr/src/linux-headers-5.10.0-3-amd64/include/generated/compile.h -rw-r--r-- 1 root root 45 Feb 6 09:23 /usr/src/linux-headers-5.10.0-3-amd64/include/generated/package.h -rw-r--r-- 1 root root 1.1K Feb 6 09:23 /usr/src/linux-headers-5.10.0-3-amd64/include/generated/timeconst.h -rw-r--r-- 1 root root 97 Feb 6 09:23 /usr/src/linux-headers-5.10.0-3-amd64/include/generated/uapi/linux/version.h -rw-r--r-- 1 root root 37 Feb 6 09:23 /usr/src/linux-headers-5.10.0-3-amd64/include/generated/utsrelease.h - Sedat - [1] https://packages.debian.org/sid/amd64/linux-headers-5.10.0-3-amd64/filelist
Bug#982304: linux-headers-5.10.0-3-amd64: File include/generated/autoconf.h missed to compile virtualbox-dkms
Package: linux-headers-5.10.0-3-amd64 Version: 5.10.13-1 Severity: normal X-Debbugs-Cc: sedat.di...@gmail.com Dear Maintainer, Today, I tried to build virtual-dkms from Debian/unstable against linux-headers-5.10.0-3-amd64. This breaks here, thus I filed Debian Bug #982301: "virtualbox-dkms: Error! Your kernel headers for kernel 5.10.0-3 cannot be found." Please, see my detailed analysis there. The important part is said in my comment #20: - BeginOfQuote - # dpkg -L linux-headers-5.10.0-3-amd64 | egrep 'auto.conf|autoconf.h' /usr/src/linux-headers-5.10.0-3-amd64/include/config/auto.conf /usr/src/linux-headers-5.10.0-3-amd64/include/generated/autoconf.h # LC_ALL=C ll /usr/src/linux-headers-5.10.0-3-amd64/include/config/auto.conf /usr/src/linux-headers-5.10.0-3-amd64/include/generated/autoconf.h ls: cannot access '/usr/src/linux-headers-5.10.0-3-amd64/include/generated/autoconf.h': No such file or directory -rw-r--r-- 1 root root 143K Feb 6 09:23 /usr/src/linux-headers-5.10.0-3-amd64/include/config/auto.conf So, the package contents list "I ship a include/generated/autoconf.h" but it does not exist :-). - EndOfQuote - Please, confirm that my analysis is correct. If yes, please provide (really) an include/generated/autoconf.h file so that DKMS is happy. Thanks. Virtual Greetings from Moenchengladbach/Germnay, - Sedat - [1] https://bugs.debian.org/982301 [2] https://bugs.debian.org/982301#20 -- System Information: Debian Release: bullseye/sid APT prefers testing APT policy: (500, 'testing'), (99, 'buildd-unstable'), (99, 'buildd-experimental'), (99, 'experimental'), (99, 'unstable') Architecture: amd64 (x86_64) Kernel: Linux 5.10.0-3-amd64 (SMP w/4 CPU threads) Locale: LANG=de_DE.UTF-8, LC_CTYPE=de_DE.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled Versions of packages linux-headers-5.10.0-3-amd64 depends on: ii linux-compiler-gcc-10-x86 5.10.13-1 ii linux-headers-5.10.0-3-common 5.10.13-1 ii linux-kbuild-5.10 5.10.13-1 linux-headers-5.10.0-3-amd64 recommends no packages. linux-headers-5.10.0-3-amd64 suggests no packages. -- no debconf information
Re: [linux-image-5.10.0-1-amd64] Requirement of pahole binary with CONFIG_DEBUG_INFO_BTF=y
On Mon, Jan 11, 2021 at 1:05 PM Sedat Dilek wrote:
>
> On Mon, Jan 11, 2021 at 12:33 PM Sedat Dilek wrote:
> >
> > On Mon, Jan 11, 2021 at 11:29 AM Salvatore Bonaccorso
> > wrote:
> > >
> > >
> > > On Mon, Jan 11, 2021 at 09:57:43AM +0100, Sedat Dilek wrote:
> > > > Hi,
> > > >
> > > > yesterday, I installed linux-image-5.10.0-1-amd64 version 5.10.5-1
> > > > from Debian/unstable.
> > > >
> > > > I am building custom Linux-kernels mostly to test with LLVM/Clang and
> > > > use Debian's kernel-config file as a base for my work.
> > > >
> > > > With CONFIG_DEBUG_INFO_BTF=y there is a requirement for pahole binary
> > > > from dwarves package.
> > > >
> > > > The requirement is coded in [1]:
> > > >
> > > > if ! [ -x "$(command -v ${PAHOLE})" ]; then
> > > > echo >&2 "BTF: ${1}: pahole (${PAHOLE}) is not available"
> > > > return 1
> > > > fi
> > > >
> > > > Is there an install-recommendation for dwarves package?
> > > > If not - I would like to open a bug-report.
> > > >
> > > > If you need further information, please let me know.
> > >
> > > this change happened in
> > > https://salsa.debian.org/kernel-team/linux/-/commit/929891281c61ce4403ddd869664c949692644a2f
> > > (actually earlier, then reverted and enabled for a subset of
> > > architectures wiith the above commit).
> > >
> > > For the Debian build it adds as well the Build-Depends on dwarves (>=
> > > 1.16~).
> > >
> > > Did I understood your question correctly?
> > >
> >
> > Hi Salvatore,
> >
> > That commit looks good to me.
> >
> > Anyway, it breaks here CONFIG_DEBUG_INFO_BTF=y:
> >
> > + [ -n y -a -n y ]
> > + info BTFIDS vmlinux
> > + [ != silent_ ]
> > + printf %-7s %s\n BTFIDS vmlinux
> > BTFIDS vmlinux
> > + ./tools/bpf/resolve_btfids/resolve_btfids vmlinux
> > FAILED: load BTF from vmlinux: Invalid argument
> > + on_exit
> > + [ 255 -ne 0 ]
> > + cleanup
> > + rm -f .btf.vmlinux.bin.o
> >
> > Any comments?
> >
>
> According to [1]:
> Q: Where do I find LLVM with BPF support?
>
> OK, I see my selfmade LLVM toolchain has no BPF support.
>
> $ which llc
> /home/dileks/src/llvm-toolchain/install/bin/llc
>
> $ llc --version
> LLVM (http://llvm.org/):
> LLVM version 11.0.1
> Optimized build.
> Default target: x86_64-unknown-linux-gnu
> Host CPU: sandybridge
>
> Registered Targets:
>x86- 32-bit X86: Pentium-Pro and above
>x86-64 - 64-bit X86: EM64T and AMD64
>
> Debian's LLC has BPF support:
>
> $ /usr/bin/llc-11 --version | grep -i bpf
>bpf- BPF (host endian)
>bpfeb - BPF (big endian)
>bpfel - BPF (little endian)
>
> I am sorry for the noise.
>
I opened an issue for tc-build (toolchain build-system for
ClangBuiltLinux folks):
"Option for BPF support when CONFIG_DEBUG_INFO_BTF=y"
- Sedat -
[1] https://github.com/ClangBuiltLinux/tc-build/issues/129
Re: [linux-image-5.10.0-1-amd64] Requirement of pahole binary with CONFIG_DEBUG_INFO_BTF=y
On Mon, Jan 11, 2021 at 12:33 PM Sedat Dilek wrote:
>
> On Mon, Jan 11, 2021 at 11:29 AM Salvatore Bonaccorso
> wrote:
> >
> >
> > On Mon, Jan 11, 2021 at 09:57:43AM +0100, Sedat Dilek wrote:
> > > Hi,
> > >
> > > yesterday, I installed linux-image-5.10.0-1-amd64 version 5.10.5-1
> > > from Debian/unstable.
> > >
> > > I am building custom Linux-kernels mostly to test with LLVM/Clang and
> > > use Debian's kernel-config file as a base for my work.
> > >
> > > With CONFIG_DEBUG_INFO_BTF=y there is a requirement for pahole binary
> > > from dwarves package.
> > >
> > > The requirement is coded in [1]:
> > >
> > > if ! [ -x "$(command -v ${PAHOLE})" ]; then
> > > echo >&2 "BTF: ${1}: pahole (${PAHOLE}) is not available"
> > > return 1
> > > fi
> > >
> > > Is there an install-recommendation for dwarves package?
> > > If not - I would like to open a bug-report.
> > >
> > > If you need further information, please let me know.
> >
> > this change happened in
> > https://salsa.debian.org/kernel-team/linux/-/commit/929891281c61ce4403ddd869664c949692644a2f
> > (actually earlier, then reverted and enabled for a subset of
> > architectures wiith the above commit).
> >
> > For the Debian build it adds as well the Build-Depends on dwarves (>=
> > 1.16~).
> >
> > Did I understood your question correctly?
> >
>
> Hi Salvatore,
>
> That commit looks good to me.
>
> Anyway, it breaks here CONFIG_DEBUG_INFO_BTF=y:
>
> + [ -n y -a -n y ]
> + info BTFIDS vmlinux
> + [ != silent_ ]
> + printf %-7s %s\n BTFIDS vmlinux
> BTFIDS vmlinux
> + ./tools/bpf/resolve_btfids/resolve_btfids vmlinux
> FAILED: load BTF from vmlinux: Invalid argument
> + on_exit
> + [ 255 -ne 0 ]
> + cleanup
> + rm -f .btf.vmlinux.bin.o
>
> Any comments?
>
According to [1]:
Q: Where do I find LLVM with BPF support?
OK, I see my selfmade LLVM toolchain has no BPF support.
$ which llc
/home/dileks/src/llvm-toolchain/install/bin/llc
$ llc --version
LLVM (http://llvm.org/):
LLVM version 11.0.1
Optimized build.
Default target: x86_64-unknown-linux-gnu
Host CPU: sandybridge
Registered Targets:
x86- 32-bit X86: Pentium-Pro and above
x86-64 - 64-bit X86: EM64T and AMD64
Debian's LLC has BPF support:
$ /usr/bin/llc-11 --version | grep -i bpf
bpf- BPF (host endian)
bpfeb - BPF (big endian)
bpfel - BPF (little endian)
I am sorry for the noise.
- Sedat -
[1]
https://www.kernel.org/doc/html/latest/bpf/bpf_devel_QA.html?highlight=pahole#llvm
Re: [linux-image-5.10.0-1-amd64] Requirement of pahole binary with CONFIG_DEBUG_INFO_BTF=y
On Mon, Jan 11, 2021 at 11:29 AM Salvatore Bonaccorso wrote:
>
>
> On Mon, Jan 11, 2021 at 09:57:43AM +0100, Sedat Dilek wrote:
> > Hi,
> >
> > yesterday, I installed linux-image-5.10.0-1-amd64 version 5.10.5-1
> > from Debian/unstable.
> >
> > I am building custom Linux-kernels mostly to test with LLVM/Clang and
> > use Debian's kernel-config file as a base for my work.
> >
> > With CONFIG_DEBUG_INFO_BTF=y there is a requirement for pahole binary
> > from dwarves package.
> >
> > The requirement is coded in [1]:
> >
> > if ! [ -x "$(command -v ${PAHOLE})" ]; then
> > echo >&2 "BTF: ${1}: pahole (${PAHOLE}) is not available"
> > return 1
> > fi
> >
> > Is there an install-recommendation for dwarves package?
> > If not - I would like to open a bug-report.
> >
> > If you need further information, please let me know.
>
> this change happened in
> https://salsa.debian.org/kernel-team/linux/-/commit/929891281c61ce4403ddd869664c949692644a2f
> (actually earlier, then reverted and enabled for a subset of
> architectures wiith the above commit).
>
> For the Debian build it adds as well the Build-Depends on dwarves (>=
> 1.16~).
>
> Did I understood your question correctly?
>
Hi Salvatore,
That commit looks good to me.
Anyway, it breaks here CONFIG_DEBUG_INFO_BTF=y:
+ [ -n y -a -n y ]
+ info BTFIDS vmlinux
+ [ != silent_ ]
+ printf %-7s %s\n BTFIDS vmlinux
BTFIDS vmlinux
+ ./tools/bpf/resolve_btfids/resolve_btfids vmlinux
FAILED: load BTF from vmlinux: Invalid argument
+ on_exit
+ [ 255 -ne 0 ]
+ cleanup
+ rm -f .btf.vmlinux.bin.o
Any comments?
Thanks.
Regards,
- Sedat -
[linux-image-5.10.0-1-amd64] Requirement of pahole binary with CONFIG_DEBUG_INFO_BTF=y
Hi,
yesterday, I installed linux-image-5.10.0-1-amd64 version 5.10.5-1
from Debian/unstable.
I am building custom Linux-kernels mostly to test with LLVM/Clang and
use Debian's kernel-config file as a base for my work.
With CONFIG_DEBUG_INFO_BTF=y there is a requirement for pahole binary
from dwarves package.
The requirement is coded in [1]:
if ! [ -x "$(command -v ${PAHOLE})" ]; then
echo >&2 "BTF: ${1}: pahole (${PAHOLE}) is not available"
return 1
fi
Is there an install-recommendation for dwarves package?
If not - I would like to open a bug-report.
If you need further information, please let me know.
Thanks.
Regards,
- Sedat -
[1]
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/scripts/link-vmlinux.sh#n144
Bug#955469: initramfs-tools-core: Enable zstandard support
Now in Linus Git tree. [1] says: Merge tag 'x86-boot-2020-08-03' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip Pull x86 boot updates from Ingo Molnar: "The main change in this cycle was to add support for ZSTD-compressed kernel and initrd images. ZSTD has a very fast decompressor, yet it compresses better than gzip" - Sedat - [1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=c0dfadfed87489fa6126ece161a14c2d15dbdc79
Bug#955469: initramfs-tools-core: Enable zstandard support
On Sat, Aug 1, 2020 at 12:27 AM Norbert Lange wrote:
>
> Am Fr., 31. Juli 2020 um 16:48 Uhr schrieb Sedat Dilek
> :
> >
> > Just FYI:
> >
> > Version 10 was now accepted in .
> >
> > Let's hope this will get into upcoming Linux v5.9.
> >
> > - Sedat -
> >
> > [1] https://github.com/terrelln/linux/commits/zstd-v10
> > [2]
> > https://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git/log/?h=x86/boot
>
> Thanks,
>
> I updated the patch, and made a merge request there:
> https://salsa.debian.org/kernel-team/initramfs-tools/-/merge_requests
>
Hi Norbert,
Thank you, too.
You mean [1]?
[ mkinitramfs ]
+zstd) compress="zstd -q -19 -T0" ;;
>From my kernel build-log:
{ cat arch/x86/boot/compressed/vmlinux.bin
arch/x86/boot/compressed/vmlinux.relocs | zstd -22 --ultra; printf
\014\015\315\001; } > arch/x86/boot/compressed/vmlinux.bin.zst
Thus I have analog here:
root# diff -uprN /usr/sbin/mkinitramfs.orig /usr/sbin/mkinitramfs
--- /usr/sbin/mkinitramfs.orig 2020-04-28 05:56:17.0 +0200
+++ /usr/sbin/mkinitramfs 2020-07-09 10:35:35.119280519 +0200
@@ -189,6 +189,7 @@ xz) compress="xz --check=crc32"
# If we're not doing a reproducible build, enable multithreading
test -z "${SOURCE_DATE_EPOCH}" && compress="$compress --threads=0"
;;
+zstd) compress="zstd -22 --ultra -v" ;;
bzip2|lzma|lzop)
# no parameters needed
;;
Regards,
- Sedat -
[1]
https://salsa.debian.org/kernel-team/initramfs-tools/-/merge_requests/33/diffs?commit_id=e94f410c71e90598b4fabca3970c7f282b5bd0a0
Bug#955469: initramfs-tools-core: Enable zstandard support
Just FYI: Version 10 was now accepted in . Let's hope this will get into upcoming Linux v5.9. - Sedat - [1] https://github.com/terrelln/linux/commits/zstd-v10 [2] https://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git/log/?h=x86/boot
linux-image-5.6.0-trunk-amd64-unsigned: Please enable CONFIG_VBOXSF_FS as module
Hi, can you please enable CONFIG_VBOXSF_FS as module? $ grep VBOX /boot/config-5.6.0-trunk-amd64 CONFIG_DRM_VBOXVIDEO=m CONFIG_VBOXGUEST=m # CONFIG_VBOXSF_FS is not set Not sure if virtualbox-guest-* like virtualbox-guest-dkms Debian packages are then co-installable and co-usable. Thanks. Regards, - Sedat - [1] https://cateee.net/lkddb/web-lkddb/VBOXSF_FS.html
diffconfig: config-5.5.0-2-amd64 VS. config-5.6.0-trunk-amd64
[ Please CC me I am not subscribed to this ML ] Hi, I installed linux-image-5.6.0-trunk-amd64-unsigned (5.6.4-1~exp1). When I compare the Kconfig changes between linux-image-5.5.0-2-amd64 (5.5.17-1) with above I see this diff: $ cd /path/to/linux $ scripts/diffconfig /boot/config-5.5.0-2-amd64 /boot/config-5.6.0-trunk-amd64 | grep ' -> ' BATMAN_ADV_SYSFS y -> n BUILD_SALT "5.5.0-2-amd64" -> "5.6.0-trunk-amd64" CEC_CORE y -> m CRC_T10DIF y -> m CRYPTO_AES y -> m CRYPTO_CBC y -> m CRYPTO_CRCT10DIF y -> m CRYPTO_CTS y -> m CRYPTO_ECB y -> m CRYPTO_LIB_AES y -> m CRYPTO_SHA512 y -> m CRYPTO_XTS y -> m ISDN_CAPI m -> y RESET_ATTACK_MITIGATION n -> y SND_HDA_PREALLOC_SIZE 2048 -> 0 SND_SOC_SOF_HDA_COMMON_HDMI_CODEC n -> y Not sure if this was intended and one of the Kconfigs is relevant for booting. Please check and let me know. Thanks. Regards, - Sedat -
Bug#955469: initramfs-tools-core: Enable zstandard support
Hi, thanks Norbert for taking care. I had a similar approach for Debian's userland tools when testing version 3 of "Add support for ZSTD-compressed kernel and initramfs" against Linux v5.6. Norbert's patch seems to include what I have in [1] (includes my README_zstd-v3.txt). I am running both ZSTD support on the kernel and userland side using a ZSTD compressed initramfs image. As a side-note: zstd-v4 is out for testing see [2] and [3]. Thanks. Regards, - Sedat - [1] LINK: https://lore.kernel.org/lkml/ca+iczuxcn2an9andrm+-enesaoygibz0w1xyhwka5k3b3u-...@mail.gmail.com/ [2] Message ID: <20200401053913.216783-1-nickrterr...@gmail.com> [3] https://github.com/terrelln/linux/commits/zstd-v4
Bug#947356: realtek-firmware: missing rtl8125a-3.fw
Hi, with linux-image-5.4.0-4-amd64 version 5.4.19-1 from Debian/testing AMD64 I can confirm this. I did this to fix it: root# cd /lib/firmware/rtl_nic root# wget https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/plain/rtl_nic/rtl8125a-3.fw NOTE: Might need to do `update-initramfs` for people having the hardware. For details see commit "rtl_nic: add firmware rtl8125a-3" [1] saying: This adds firmware rtl8125a-3 for Realtek's 2.5Gbps chip RTL8125. Thanks. Regards, - Sedat - [1] https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/commit/rtl_nic?id=f667c005600bd4fe24a0a439b7a3f3eadcce753a
Bug#902966: pstore: crypto_comp_decompress failed
Hi, I can confirm to see this on Debian/buster AMD64 with... linux-image-4.17.0-1-amd64 (4.17.8-1) ...and removing all dmesg-efi-*.enc.z files in /sys/fs/pstore/... ...and rebooting made the failures in dmesg go away. Not sure what 'gzip: stdin: unexpected end of file' means exactly in this context (see below). I can send all relevant files if needed. Thanks, - Sedat - P.S.: CSI pstore-deflate failures root# egrep -i 'pstore|deflate' /boot/config-4.17.0-1-amd64 CONFIG_PPP_DEFLATE=m CONFIG_CHROMEOS_PSTORE=m CONFIG_EFI_VARS_PSTORE=m # CONFIG_EFI_VARS_PSTORE_DEFAULT_DISABLE is not set CONFIG_PSTORE=y CONFIG_PSTORE_DEFLATE_COMPRESS=y # CONFIG_PSTORE_LZO_COMPRESS is not set # CONFIG_PSTORE_LZ4_COMPRESS is not set # CONFIG_PSTORE_LZ4HC_COMPRESS is not set # CONFIG_PSTORE_842_COMPRESS is not set CONFIG_PSTORE_COMPRESS=y CONFIG_PSTORE_DEFLATE_COMPRESS_DEFAULT=y CONFIG_PSTORE_COMPRESS_DEFAULT="deflate" # CONFIG_PSTORE_CONSOLE is not set # CONFIG_PSTORE_PMSG is not set # CONFIG_PSTORE_FTRACE is not set CONFIG_PSTORE_RAM=m CONFIG_CRYPTO_DEFLATE=y CONFIG_ZLIB_DEFLATE=y root# printf "\x1f\x8b\x08\x00\x00\x00\x00\x00" | cat - dmesg-efi-152896783215001.enc.z | gzip -dc Oops#1 Part15 <6>[2.324428] async_tx: api initialized (async) <6>[2.338058] device-mapper: uevent: version 1.0.3 <6>[2.339096] device-mapper: ioctl: 4.37.0-ioctl (2017-09-20) initialised: dm-de...@redhat.com <5>[2.366740] random: fast init done <6>[2.416909] usb 1-2: New USB device found, idVendor=046d, idProduct=c077 <6>[2.416913] usb 1-2: New USB device strings: Mfr=1, Product=2, SerialNumber=0 <6>[2.416916] usb 1-2: Product: USB Optical Mouse <6>[2.416918] usb 1-2: Manufacturer: Logitech <6>[2.419553] input: Logitech USB Optical Mouse as /devices/pci:00/:00:14.0/usb1/1-2/1-2:1.0/0003:046D:C077.0003/input/input5 <6>[2.419852] hid-generic 0003:046D:C077.0003: input,hidraw2: USB HID v1.11 Mouse [Logitech USB Optical Mouse] on usb-:00:14.0-2/input0 <6>[2.548486] usb 1-3: new full-speed USB device number 4 using xhci_hcd <6>[2.684147] psmouse serio1: synaptics: queried max coordinates: x [..5676], y [..4690] <6>[2.707942] usb 1-3: New USB device found, idVendor=058f, idProduct=9540 <6>[2.707947] usb 1-3: New USB device strings: Mfr=1, Product=2, SerialNumber=0 <6>[2.707951] usb 1-3: Product: EMV Smartcard Reader <6>[2.707954] usb 1-3: Manufacturer: Generic <6>[2.720069] psmouse serio1: synaptics: queried min coordinates: x [1266..], y [1162..] <4>[2.720079] psmouse serio1: synaptics: The touchpad can support a better bus than the too old PS/2 protocol. Make sure MOUSE_PS2_SYNAPTICS_SMBUS and RMI4_SMB are enabled to get a better touchpad experience. <6>[2.792567] psmouse serio1: synaptics: Touchpad model: 1, fw: 8.2, id: 0x1e2b1, caps: 0xf002a3/0x940300/0x12e800/0x40, board id: 3276, fw id: 2491725 <6>[2.792582] psmouse serio1: synaptics: serio: Synaptics pass-through port at isa0060/serio1/input0 gzip: stdin: unexpected end of file - EOT -
Re: dkms: Build with an alternative like compiler rather than gcc
On Mon, Jul 9, 2018 at 4:43 PM, wrote: >> -Original Message- >> From: Sedat Dilek [mailto:sedat.di...@gmail.com] >> Sent: Monday, July 9, 2018 9:41 AM >> To: Limonciello, Mario >> Cc: p...@hungry.com; da...@debian.org; a...@debian.org; >> iucul...@debian.org; pkg-dkms-ma...@lists.alioth.debian.org; debian- >> ker...@lists.debian.org >> Subject: Re: dkms: Build with an alternative like compiler rather than gcc >> >> On Mon, Jul 9, 2018 at 4:37 PM, wrote: >> >> -Original Message- >> >> From: Petter Reinholdtsen [mailto:p...@hungry.com] >> >> Sent: Monday, July 9, 2018 9:32 AM >> >> To: sedat.di...@gmail.com; David Paleino; Aron Xu; Limonciello, Mario; >> Giuseppe >> >> Iuculano >> >> Cc: Dynamic Kernel Modules Support Team; debian-kernel@lists.debian.org >> >> Subject: Re: dkms: Build with an alternative like compiler rather than gcc >> >> >> >> >> >> [Sedat Dilek] >> >> > As a workaround I have symlinked my mycompiler wrapper-script to >> >> > /usr/bin/gcc. That works. >> >> > >> >> > What is the recommended way to do this correctly? >> >> >> >> Perhaps you should use dpkg-divert to ensure your replaced gcc is not >> >> lost in a future package upgrade? >> >> >> >> The DKMS team in Debian need active members. I'm not one of them. :( >> >> >> >> -- >> > >> > I think this is an upstream problem not at all specific to Debian's >> > implementation. >> > I would recommend to bring this discussion upstream to get it fixed there. >> > >> > I believe DKMS unsets CC before starting build, but maybe that's wrong to >> > do >> > in this instance. >> > >> >> Personally, I do not think it is an upstream problem. >> Unsetting CC before DKMS starts the build would explain the >> behaviour/handling. >> > [ Removing wrong debian's dkms maintainer list ] > How exactly does clang get used by kernel build? It's by setting CC variable > right? > You can try to remove this line to see if it fixes the problem (it should). > https://github.com/dell/dkms/blob/master/dkms#L3515 > I have an own script which uses bindeb-pkg make-target from Linus tree. OK, that line looks like the culprit. Anyway, I need a flexible handling when using different compilers - for me. > The question would be where that variable should be ignored and where it > shouldn't. > Which I believe is an upstream discussion not a Debian discussion. I agree. - sed@ -
Re: dkms: Build with an alternative like compiler rather than gcc
On Mon, Jul 9, 2018 at 4:37 PM, wrote: >> -Original Message- >> From: Petter Reinholdtsen [mailto:p...@hungry.com] >> Sent: Monday, July 9, 2018 9:32 AM >> To: sedat.di...@gmail.com; David Paleino; Aron Xu; Limonciello, Mario; >> Giuseppe >> Iuculano >> Cc: Dynamic Kernel Modules Support Team; debian-kernel@lists.debian.org >> Subject: Re: dkms: Build with an alternative like compiler rather than gcc >> >> >> [Sedat Dilek] >> > As a workaround I have symlinked my mycompiler wrapper-script to >> > /usr/bin/gcc. That works. >> > >> > What is the recommended way to do this correctly? >> >> Perhaps you should use dpkg-divert to ensure your replaced gcc is not >> lost in a future package upgrade? >> >> The DKMS team in Debian need active members. I'm not one of them. :( >> >> -- > > I think this is an upstream problem not at all specific to Debian's > implementation. > I would recommend to bring this discussion upstream to get it fixed there. > > I believe DKMS unsets CC before starting build, but maybe that's wrong to do > in this instance. > Personally, I do not think it is an upstream problem. Unsetting CC before DKMS starts the build would explain the behaviour/handling. - sed@ -
Re: dkms: Build with an alternative like compiler rather than gcc
On Mon, Jul 9, 2018 at 4:31 PM, Petter Reinholdtsen wrote: > > [Sedat Dilek] >> As a workaround I have symlinked my mycompiler wrapper-script to >> /usr/bin/gcc. That works. >> >> What is the recommended way to do this correctly? > > Perhaps you should use dpkg-divert to ensure your replaced gcc is not > lost in a future package upgrade? > > The DKMS team in Debian need active members. I'm not one of them. :( > I am thinking of using mycompiler as an cc alternative, the same with c++ and cpp. The wrapper-script(s) let me pass compiler-options, too. So this is very flexible. Active members... the alioth email-adress in the dsc-file seems no longer valid. Thanks for the response. - sed@ -
dkms: Build with an alternative like compiler rather than gcc
Hi, I am experimenting with clang to compile an upstream Linux kernel v4.14.y (here: y=54). For this I use bindeb-pkg make-target from Linus upstream. After installing the generated linux-headers and linux-image amd64 packages, the dkms kernel-modules were built with gcc-7 which defaults to gcc. As a result acpi-call and virtualbox dkms modules do not load (see [1]). As a workaround I have symlinked my mycompiler wrapper-script to /usr/bin/gcc. That works. What is the recommended way to do this correctly? Is there an CC option for dkms.conf? Use the update-alternatives system for configuring/installing clang as an cc-alternative? Thanks. Regards, - Sedat - [1] https://lkml.org/lkml/2018/6/1/724 [2] https://stackoverflow.com/questions/7832892/how-to-change-the-default-gcc-compiler-in-ubuntu P.S.: Some instructions and informations which might be helpful root@iniza# LANG=C update-alternatives --display cc cc - auto mode link best version is /usr/bin/gcc link currently points to /usr/bin/gcc link cc is /usr/bin/cc /usr/bin/gcc - priority 20 root# dpkg --purge tp-smapi-dkms <--- XXX: Does not work with Lenovo ThinkPad T470 root# dkms status -k 4.14.54-1-iniza-llvmlinux acpi-call, 1.1.0, 4.14.54-1-iniza-llvmlinux, x86_64: installed virtualbox, 5.2.10, 4.14.54-1-iniza-llvmlinux, x86_64: installed root# dkms remove acpi-call -v 1.1.0 -k 4.14.54-1-iniza-llvmlinux root# dkms remove virtualbox -v 5.2.10 -k 4.14.54-1-iniza-llvmlinux root# cat /usr/bin/mycompiler #!/bin/bash ccache clang-7 "$@" root# cd /usr/bin ; ln -sf mycompiler gcc <--- XXX: Workaround root# LANG=C dkms install acpi-call -v 1.1.0 -k 4.14.54-1-iniza-llvmlinux root# LANG=C dkms install virtualbox -v 5.2.10 -k 4.14.54-1-iniza-llvmlinux root# cd /usr/bin ; ln -sf gcc-7 gcc <--- XXX: System default
[linux-stable-3.16.y] tun: allow positive return values on dev_get_valid_name() call
Hi Ben, some Debian/jessie systems were caught by the bug-report in [1]. This issue was recently fixed in an updated Debian kernel for v3.16.y. Will you include the patch "tun: allow positive return values on dev_get_valid_name() call" [2] in linux-stable-3.16.y upstream? This was a fix for "tun: call dev_get_valid_name() before register_netdevice()" [3]. Unfortunately, there is not a reference (usually "Fixes:" tag) for this in [2]. Not sure if this was documented in the meantime like [4] says. Thanks in advance. Regards, - Sedat - [1] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=897427 [2] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=5c25f65fd1e42685f7ccd80e0621829c105785d9 [3] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=0ad646c81b2182f7fa67ec0c8c825e0ee165696d [4] https://www.spinics.net/lists/netdev/msg462705.html
Bug#897685: linux-image-3.16.0-6-amd64: Unable to start multiple KVM instances with libvirt
Hi, yesterday, we had maintenance day and wanted to upgrade our Ganeti clusters. Our Ganeti clusters use KVM and libvirt. Clusters running with Debian/stretch were not affected whereas clusters running Debian/jessie failed reproducible when starting more than 2 Ganeti instances. First, we want to confirm this issue with linux-image-3.16.0-6-amd64. We saw DRBD errors in our logs like... drbdX: Discarding network configuration (X = Number of Ganeti instance) As a workaround, we did a downgrade to the previous Debian-kernel release and put it on hold: root# echo linux-image-3.16.0-5-amd64 hold | dpkg --set-selections root# echo linux-image-amd64 hold | dpkg --set-selections root# dpkg --get-selections | grep hold linux-image-3.16.0-5-amd64 hold linux-image-amd64 hold Hope this helps. Thanks, - Sedat - [1] https://www.mail-archive.com/debian-bugs-dist@lists.debian.org/msg1602078.html [2] https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git/log/?h=v3.16.56 -- Mit freundlichen Grüssen, Sedat Dilek Telefon: +49 2166 9901-153 E-Mail: sedat.di...@credativ.de Internet: https://www.credativ.de/ GPG-Fingerprint: EA6D E17D D269 AC7E 101D C910 476F 2B3B 0AF7 F86B credativ GmbH, Trompeterallee 108, 41189 Mönchengladbach Handelsregister: Amtsgericht Mönchengladbach, HR-Nummer: HRB 12080, UID-Nummer: DE204566209 Geschäftsführung: Dr. Michael Meskes, Jörg Folz, Sascha Heuer ** Jetzt neu: Elephant Shed - PostgreSQL Appliance PostgreSQL und alles was dazugehört Von Backup über Monitoring bis Reporting: https://elephant-shed.io/ **
[Best practice] Testing custom patchset
Hi,
I wanted to build a Debian-kernel by applying Linux v4.14.13-rc1
patchset against linux (4.14.12-2) from Debian/unstable.
My system is a Debian/stretch AMD64.
I consulted the Debian-kernel handbook [1] to do this the "Debian way".
The patchset was extraxted as a series from
linux-stable-rc.git#linux-4.14.y and placed in an "upstream" directory
below debian/patches...
cd debian/patches
ls upstream/*.patch >> series
I commented "already applied-in-debian-kernel" or unwanted patches
(see below P.S.).
As I downgraded to use gcc-6 as default-compiler (gcc-7 is not
available in stretch, see [2]) and changed abiname, I executed...
./debian/bin/gencontrol.py
For a fast build I want to build a single-kernel-variant (here:
amd64_non_amd64)
ARCH="amd64"
FEATURESET="none"
FLAVOUR="amd64"
SINGLE_KERNEL_VARIANT="${ARCH}_${FEATURESET}_${FLAVOUR}"
fakeroot make -f debian/rules.gen setup_$SINGLE_KERNEL_VARIANT
[1] recommends to use test-patches script...
LC_ALL=C bash ./debian/bin/test-patches -f amd64 -s none debian/patches/series
I couldn't invoke the script without passing -f and -s (none seems to
be default/fallback).
This is applying the patchset from debian/patches/series-rt.
Am I doing something wrong?
Is something wrong with the script?
As an alternative I tried...
LC_ALL=C debuild -S
...to see if the patches apply, they do.
Building linux-image and linux-headers-arch Debian packages...
MAX_JOBS=$(($(getconf _NPROCESSORS_ONLN)+1))
PARALLEL_MAKE_JOBS="$MAX_JOBS"
fakeroot make -j${PARALLEL_MAKE_JOBS} -f debian/rules.gen
binary-arch_$SINGLE_KERNEL_VARIANT
Generating missing linux-headers-common Debian package...
fakeroot make -f debian/rules.gen binary-indep_${FEATURESET}
Can you give me advices on how to test (and apply) a custom patchset
before building?
Any recommendation where to place it (existing directories, own
directory) and embed in an existing or own series file?
If you need furher informations, please let me know.
I can send you my dsc and debian-tarball is needed.
Thanks in advance.
Regards,
- Sedat -
[1] https://kernel-handbook.alioth.debian.org/ch-common-tasks.html
[2]
https://anonscm.debian.org/cgit/kernel/linux.git/patch/?id=181f3cd32b3d76d75813a62b230c0ed0fd7e0812
P.S.: Extract of debian/patches/series
...
# Upstream patches from Linux-4.14.13-rc1
upstream/0001-x86-mm-Set-MODULES_END-to-0xff00.patch
upstream/0002-x86-mm-Map-cpu_entry_area-at-the-same-place-on-4-5-l.patch
upstream/0003-x86-kaslr-Fix-the-vaddr_end-mess.patch
upstream/0004-x86-events-intel-ds-Use-the-proper-cache-flush-metho.patch
##upstream/0005-x86-tlb-Drop-the-_GPL-from-the-cpu_tlbstate-export.patch
upstream/0006-x86-alternatives-Add-missing-n-at-end-of-ALTERNATIVE.patch
upstream/0007-x86-pti-Rename-BUG_CPU_INSECURE-to-BUG_CPU_MELTDOWN.patch
upstream/0008-kernel-acct.c-fix-the-acct-needcheck-check-in-check_.patch
upstream/0009-mm-mprotect-add-a-cond_resched-inside-change_pmd_ran.patch
upstream/0010-mm-sparse.c-wrong-allocation-for-mem_section.patch
upstream/0011-userfaultfd-clear-the-vma-vm_userfaultfd_ctx-if-UFFD.patch
upstream/0012-btrfs-fix-refcount_t-usage-when-deleting-btrfs_delay.patch
upstream/0013-efi-capsule-loader-Reinstate-virtual-capsule-mapping.patch
upstream/0014-crypto-n2-cure-use-after-free.patch
upstream/0015-crypto-chacha20poly1305-validate-the-digest-size.patch
upstream/0016-crypto-pcrypt-fix-freeing-pcrypt-instances.patch
upstream/0017-crypto-chelsio-select-CRYPTO_GF128MUL.patch
upstream/0018-drm-i915-Disable-DC-states-around-GMBUS-on-GLK.patch
upstream/0019-drm-i915-Apply-Display-WA-1183-on-skl-kbl-and-cfl.patch
upstream/0020-sunxi-rsb-Include-OF-based-modalias-in-device-uevent.patch
upstream/0021-fscache-Fix-the-default-for-fscache_maybe_release_pa.patch
upstream/0022-x86-CPU-Avoid-unnecessary-IPIs-in-arch_freq_get_on_c.patch
upstream/0023-x86-CPU-Always-show-current-CPU-frequency-in-proc-cp.patch
upstream/0024-kernel-signal.c-protect-the-traced-SIGNAL_UNKILLABLE.patch
upstream/0025-kernel-signal.c-protect-the-SIGNAL_UNKILLABLE-tasks-.patch
upstream/0026-kernel-signal.c-remove-the-no-longer-needed-SIGNAL_U.patch
upstream/0027-iommu-arm-smmu-v3-Don-t-free-page-table-ops-twice.patch
upstream/0028-iommu-arm-smmu-v3-Cope-with-duplicated-Stream-IDs.patch
upstream/0029-ARC-uaccess-dont-use-l-gcc-inline-asm-constraint-mod.patch
upstream/0030-powerpc-mm-Fix-SEGV-on-mapped-region-to-return-SEGV_.patch
upstream/0031-Input-elantech-add-new-icbody-type-15.patch
upstream/0032-x86-microcode-AMD-Add-support-for-fam17h-microcode-l.patch
upstream/0033-apparmor-fix-regression-in-mount-mediation-when-feat.patch
upstream/0034-parisc-Fix-alignment-of-pa_tlb_lock-in-assembly-on-3.patch
upstream/0035-parisc-qemu-idle-sleep-support.patch
upstream/0036-mtd-nand-pxa3xx-Fix-READOOB-implementation.patch
upstream/0037-KVM-s390-fix-cmma-migration-for-multiple-memory-slot.patch
upstream/0038-KVM-s390-prevent-buffer-overrun-on-memory-hotplug-du.patch
##upstream/0039-Linux-4.14.13-rc1.patch
-
Re: [debian/stretch64] lxc-copy: Snapshot with OverlayFS backingstorage fails with Linux-4.9.y
On Wed, Nov 1, 2017 at 2:18 PM, Christian Brauner <christian.brau...@mailbox.org> wrote: > On Wed, Oct 25, 2017 at 11:45:52AM +0200, Sedat Dilek wrote: >> [ Please CC me - I am not subscribed to lxc-users and debian-kernel MLs ] >> >> Hi, >> >> I am playing with LXC and OverlayFS. >> >> The Debian-kernel shipped with stretch64 fails when using lxc-copy for >> a snapshot. >> Later I tried the latest Debian-kernel from stretch-backports >> APT-repository [0] and the issue does *not* appear. >> >> Debian/stretch64 ships LXC v2.0.7. >> >> root# lxc-copy --version >> 2.0.7 >> >> [ Instructions to reproduce ] >> >> [ Create LXC-Container ] >> >> root# lxc-create -n stretch64-ovl -B overlayfs -t download -- -d >> debian -r stretch -a amd64 >> >> [ GOOD: linux-image-4.12.0-0.bpo.2-amd64 (4.12.13-1~bpo9+1) ] >> >> root# lxc-copy -n stretch64-ovl -N snapshot1 -B overlay -s >> --logfile=lxc-copy-log_kernel-bpo9.txt --logpriority=DEBUG >> >> [ BAD: linux-image-4.9.0-4-amd64 (4.9.51-1) ] >> >> root@iniza:~# lxc-copy -n stretch64-ovl -N snapshot2 -B overlay -s >> --logfile=lxc-copy-log_kernel-deb9.txt --logpriority=DEBUG >> >> I found lxc-issue #1395 [1] where a similiar problem was reported >> >> Can this be a kernel issue? > > Well, if the only different in your tests is the kernel then could be. But > there > also have been quite some changes to the liblxc codebase that should improve > general user experience wrt to the overlay storage driver. So you might want > to > test e.g. liblxc 2.1. If your environment allows you to do that. > [ CC Evegeni (lxc debian maintainer team) ] Hi Christian, The background is a student project where we use Debian/stretch as a base for the lxc-containers. So upgrading liblxc is problematic. I asked Evgeni for a stretch-backport package of v2.0.9 - I can test with that and report when the package is available. The issue in [1] is similiar and has status "incomplete"? What was missing? How can I help? Thanks, - Sedat - [1] https://github.com/lxc/lxc/issues/1395 >> >> Miklos and Christian any idea? >> >> Do you need further informations and/or have instructions on debugging this? >> >> Thanks for any hints. >> >> Regards, >> - Sedat - >> >> P.S.: I have attached the LXC configs of stretch64-ovl & snaptshot1 >> and the lxc-copy logfiles. >> >> [0] https://backports.debian.org/Instructions/ >> [1] https://github.com/lxc/lxc/issues/1395 > >>lxc-copy 20171025093206.756 WARN lxc_confile - >> confile.c:config_pivotdir:1910 - lxc.pivotdir is ignored. It will soon >> become an error. >>lxc-copy 20171025093206.756 WARN lxc_confile - >> confile.c:config_pivotdir:1910 - lxc.pivotdir is ignored. It will soon >> become an error. >>lxc-copy 20171025093206.756 INFO lxcoverlay - >> bdev/lxcoverlay.c:ovl_mount:436 - Overlayfs: Mounted >> /var/lib/lxc/stretch64-ovl/rootfs onto /var/lib/lxc/stretch64-ovl/rootfs >> with options >> upperdir=/var/lib/lxc/stretch64-ovl/delta0,lowerdir=/var/lib/lxc/stretch64-ovl/rootfs,workdir=/var/lib/lxc/stretch64-ovl/olwork. >>lxc-copy 20171025093206.757 INFO lxcoverlay - >> bdev/lxcoverlay.c:ovl_mount:436 - Overlayfs: Mounted >> /var/lib/lxc/stretch64-ovl/rootfs onto /var/lib/lxc/snapshot2/rootfs with >> options >> upperdir=/var/lib/lxc/snapshot2/delta0,lowerdir=/var/lib/lxc/stretch64-ovl/rootfs,workdir=/var/lib/lxc/snapshot2/olwork. >>lxc-copy 20171025093207.420 ERRORlxcoverlay - >> bdev/lxcoverlay.c:ovl_rsync:717 - rsyncing /var/lib/lxc/stretch64-ovl/rootfs >> to /var/lib/lxc/snapshot2/rootfs >>lxc-copy 20171025093207.421 ERRORlxcoverlay - >> bdev/lxcoverlay.c:ovl_do_rsync:757 - copying overlayfs delta >>lxc-copy 20171025093207.421 ERRORbdev - bdev/bdev.c:bdev_copy:388 >> - failed getting pathnames for cloned storage: >> overlayfs:/var/lib/lxc/stretch64-ovl/rootfs:/var/lib/lxc/stretch64-ovl/delta0 >>lxc-copy 20171025093207.422 ERRORlxc_container - >> lxccontainer.c:copy_storage:2867 - Error copying storage. >>lxc-copy 20171025093207.430 INFO lxc_container - >> lxccontainer.c:container_destroy:2414 - Destroyed directory for snapshot2 > > > >>lxc-copy 20171025091007.914 WARN lxc_confile - >> confile.c:config_pivotdir:1910 - lxc.pivotdir is ignored. It will soon >> become an error. >>lxc-copy 20171025091007.914 WARN lxc_confile - >> confile.c:config_pivo
[debian/stretch64] lxc-copy: Snapshot with OverlayFS backingstorage fails with Linux-4.9.y
[ Please CC me - I am not subscribed to lxc-users and debian-kernel MLs ] Hi, I am playing with LXC and OverlayFS. The Debian-kernel shipped with stretch64 fails when using lxc-copy for a snapshot. Later I tried the latest Debian-kernel from stretch-backports APT-repository [0] and the issue does *not* appear. Debian/stretch64 ships LXC v2.0.7. root# lxc-copy --version 2.0.7 [ Instructions to reproduce ] [ Create LXC-Container ] root# lxc-create -n stretch64-ovl -B overlayfs -t download -- -d debian -r stretch -a amd64 [ GOOD: linux-image-4.12.0-0.bpo.2-amd64 (4.12.13-1~bpo9+1) ] root# lxc-copy -n stretch64-ovl -N snapshot1 -B overlay -s --logfile=lxc-copy-log_kernel-bpo9.txt --logpriority=DEBUG [ BAD: linux-image-4.9.0-4-amd64 (4.9.51-1) ] root@iniza:~# lxc-copy -n stretch64-ovl -N snapshot2 -B overlay -s --logfile=lxc-copy-log_kernel-deb9.txt --logpriority=DEBUG I found lxc-issue #1395 [1] where a similiar problem was reported Can this be a kernel issue? Miklos and Christian any idea? Do you need further informations and/or have instructions on debugging this? Thanks for any hints. Regards, - Sedat - P.S.: I have attached the LXC configs of stretch64-ovl & snaptshot1 and the lxc-copy logfiles. [0] https://backports.debian.org/Instructions/ [1] https://github.com/lxc/lxc/issues/1395 lxc-copy 20171025093206.756 WARN lxc_confile - confile.c:config_pivotdir:1910 - lxc.pivotdir is ignored. It will soon become an error. lxc-copy 20171025093206.756 WARN lxc_confile - confile.c:config_pivotdir:1910 - lxc.pivotdir is ignored. It will soon become an error. lxc-copy 20171025093206.756 INFO lxcoverlay - bdev/lxcoverlay.c:ovl_mount:436 - Overlayfs: Mounted /var/lib/lxc/stretch64-ovl/rootfs onto /var/lib/lxc/stretch64-ovl/rootfs with options upperdir=/var/lib/lxc/stretch64-ovl/delta0,lowerdir=/var/lib/lxc/stretch64-ovl/rootfs,workdir=/var/lib/lxc/stretch64-ovl/olwork. lxc-copy 20171025093206.757 INFO lxcoverlay - bdev/lxcoverlay.c:ovl_mount:436 - Overlayfs: Mounted /var/lib/lxc/stretch64-ovl/rootfs onto /var/lib/lxc/snapshot2/rootfs with options upperdir=/var/lib/lxc/snapshot2/delta0,lowerdir=/var/lib/lxc/stretch64-ovl/rootfs,workdir=/var/lib/lxc/snapshot2/olwork. lxc-copy 20171025093207.420 ERRORlxcoverlay - bdev/lxcoverlay.c:ovl_rsync:717 - rsyncing /var/lib/lxc/stretch64-ovl/rootfs to /var/lib/lxc/snapshot2/rootfs lxc-copy 20171025093207.421 ERRORlxcoverlay - bdev/lxcoverlay.c:ovl_do_rsync:757 - copying overlayfs delta lxc-copy 20171025093207.421 ERRORbdev - bdev/bdev.c:bdev_copy:388 - failed getting pathnames for cloned storage: overlayfs:/var/lib/lxc/stretch64-ovl/rootfs:/var/lib/lxc/stretch64-ovl/delta0 lxc-copy 20171025093207.422 ERRORlxc_container - lxccontainer.c:copy_storage:2867 - Error copying storage. lxc-copy 20171025093207.430 INFO lxc_container - lxccontainer.c:container_destroy:2414 - Destroyed directory for snapshot2 config_snapshot1 Description: Binary data config_stretch64-ovl Description: Binary data lxc-copy 20171025091007.914 WARN lxc_confile - confile.c:config_pivotdir:1910 - lxc.pivotdir is ignored. It will soon become an error. lxc-copy 20171025091007.914 WARN lxc_confile - confile.c:config_pivotdir:1910 - lxc.pivotdir is ignored. It will soon become an error. lxc-copy 20171025091007.914 INFO lxcoverlay - bdev/lxcoverlay.c:ovl_mount:436 - Overlayfs: Mounted /var/lib/lxc/stretch64-ovl/rootfs onto /var/lib/lxc/stretch64-ovl/rootfs with options upperdir=/var/lib/lxc/stretch64-ovl/delta0,lowerdir=/var/lib/lxc/stretch64-ovl/rootfs,workdir=/var/lib/lxc/stretch64-ovl/olwork. lxc-copy 20171025091007.914 INFO lxcoverlay - bdev/lxcoverlay.c:ovl_mount:436 - Overlayfs: Mounted /var/lib/lxc/stretch64-ovl/rootfs onto /var/lib/lxc/snapshot1/rootfs with options upperdir=/var/lib/lxc/snapshot1/delta0,lowerdir=/var/lib/lxc/stretch64-ovl/rootfs,workdir=/var/lib/lxc/snapshot1/olwork. lxc-copy 20171025091008.114 INFO lxc_container - lxccontainer.c:copy_file:2641 - Error stat'ing /var/lib/lxc/stretch64-ovl/lxc_rdepends lxc-copy 20171025091008.114 INFO lxc_container - lxccontainer.c:copy_rdepends:2800 - Error copying reverse dependencies lxc-copy 20171025091008.136 INFO lxcoverlay - bdev/lxcoverlay.c:ovl_mount:436 - Overlayfs: Mounted /var/lib/lxc/stretch64-ovl/rootfs onto /usr/lib/x86_64-linux-gnu/lxc/rootfs with options upperdir=/var/lib/lxc/snapshot1/delta0,lowerdir=/var/lib/lxc/stretch64-ovl/rootfs,workdir=/var/lib/lxc/snapshot1/olwork. lxc-copy 20171025091008.136 INFO lxc_conf - conf.c:run_script_argv:424 - Executing script "/usr/share/lxc/hooks/clonehostname" for container "snapshot1", config section "lxc". lxc-copy 20171025091008.166 INFO lxc_copy_ui - tools/lxc_copy.c:do_clone:376 - Created snapshot1 as snapshot of stretch64-ovl
Re: linux-image-3.18.0-trunk-amd64 (3.18.5-1~exp1)
Hi, I am comparing my own kernel-config with the ones from Ubuntu and Debian. Just curious for what 0 stands in 3.18.0-trunk? ABIVER? It's reflecting v3.18.y where y=5 and it should be 3.18.5-trunk-amd64? Thanks for clarification. Regards, - Sedat - -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/ca+iczuvaefbs7hemgvapqn33fspgtjslenvnk13q-rhkxmu...@mail.gmail.com
Bug#759809: linux-image-3.2.0-4-amd64: USB 3 devices fail, logging xHCI xhci_drop_endpoint called with disabled ep ffff8802165db0c0
On Wed, Nov 5, 2014 at 10:19 PM, Henrik Størner hen...@hswn.dk wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 01-11-2014 kl. 19:07 Ben Hutchings wrote: On Sat, 2014-11-01 at 14:56 +0100, Sedat Dilek wrote: On Tue, Sep 9, 2014 at 4:03 PM, Sedat Dilek sedat.di...@gmail.com wrote: Hi, this looks like what I reported in [wheezy][linux-3.2.60] Booting from a Debian system on an external USB-3.0 hdd hangs (see [1]). Unfortunately, there is no newer 3.2.y kernel for wheezy available [2]. Henrik, did you try a Linux-kernel = 3.2.61 (self-compiled)? Thanks. Regards, - Sedat - [1] https://lists.debian.org/debian-kernel/2014/08/msg00141.html [2] http://anonscm.debian.org/viewvc/kernel/dists/wheezy/linux/ The problem was solved by updating to linux-image-3.2.0-4-amd64 (3.2.63-2+deb7u1). Henrik, can you confirm this is also fixed on your computer? Confirmed, upgraded to 3.2.63-2+deb7u1 and my disk works again using the xhci_hcd driver. Thanks :-) Will that BR be closed now? - Sedat - Regards, Henrik -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.12 (GNU/Linux) iQEVAwUBVFqUSPADgvSOCWu5AQJ5fgf8CJB1SGP1g3CbdxldVGfCU4nFBLrMJpE+ mrxOiRzC13KSWrRVAfyWNgkaO5L+7UtiCNYu6qKo6wFmY74rxuPjYX0cQAu0LsP+ v0Bz6PyxZQGi0HMIvnutiRJAIvBi7TA0xa3W16QfL81dIvcQi+JTr/WSP7iin+6O wjX5eNVm84rlXxFP2OTBxR5AF8KQZSeThoLS5mtP9E0sKhorq+2nnjPQHxmtWMGq Hd8TMea+dAdSskFVA0EUbLj8mzVuzhMI2kaZvUezCNRiShAiOiK2u6Vr8Y6P5wOu tHAZOISwPk+uFU3YxeE7uy9IESq869rFWLGUumjjuv9nYXzJZdGlYw== =OQrN -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/ca+iczuw+y0fweceil_ecodwywmtms7oz8+ngyicem9ksn...@mail.gmail.com
Bug#759809: linux-image-3.2.0-4-amd64: USB 3 devices fail, logging xHCI xhci_drop_endpoint called with disabled ep ffff8802165db0c0
On Sat, Nov 1, 2014 at 3:16 PM, Sedat Dilek sedat.di...@gmail.com wrote: On Sat, Nov 1, 2014 at 2:56 PM, Sedat Dilek sedat.di...@gmail.com wrote: [...] The problem was solved by updating to linux-image-3.2.0-4-amd64 (3.2.63-2+deb7u1). Just for the sake of completeness (and for Jari)... Here I have an ASM1042 chipset... Looks like that chip has also some troubles with UAS (see [1] xhci: Disable streams on Asmedia 1042 xhci controllers). Not sure if Linux-3.2.y has the xhci-quirks-handling implemented. - Sedat - [1] http://git.kernel.org/cgit/linux/kernel/git/gregkh/usb.git/commit/?h=usb-linusid=2391eacbd00b706ff4902db7dbee21e33b6f1850 -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/CA+icZUWDXGEyM65POP3HYXm=cnmfa_0rz1bt9gnksr4ndxo...@mail.gmail.com
Re: [wheezy][linux-3.2.60] Booting from a Debian system on an external USB-3.0 hdd hangs
On Tue, Aug 12, 2014 at 8:16 AM, Sedat Dilek sedat.di...@gmail.com wrote: [ Please CC me - I am not subscribed to this ML ] Hi, I cannot boot with my current linux-3.2.y Debian/wheezy kernel as my system hangs and reports xhci problems. ( Unfortunately, nothing in the logs and no picture from my screen for you ). FYI: My system and bootloader etc. is on an external USB-3.0 (xhci) hdd. linux-image (3.2.57-3+deb7u1) was OK and (3.2.60-1+deb7u1) was 1st causing the problem. AFAICS this is the missing fix which is included in linux-3.2.61 [1]. commit 5dc2808c4729bf080487e61b80ee04e0fdb12a37 upstream. xhci: delete endpoints from bandwidth list before freeing whole device I have seen linux-3.2.62 was released recently. Can you integrate above fix as a hotfix or do you plan to push a linux-3.2.62 kernel soon? Thanks in advance. Regards, - Sedat - P.S.: If you need additional informations, please let me know. [1] http://git.kernel.org/cgit/linux/kernel/git/bwh/linux-3.2.y-queue.git /tree/releases/3.2.61/xhci-delete-endpoints-from-bandwidth-list-befor e-freeing-whole.patch Just a short feedback... After I had rescued my Debian/wheezy in a chroot-env [1] I can boot into it with linux-image-3.2.0-4-amd64 (3.2.63-2+deb7u1) again. - Sedat - [1] http://manual.aptosid.com/en/sys-admin-grub2-en.htm#chroot P.S.: [1] misses a note to copy the local /etc/resolv.conf to the chroot-env in case of networking problems. -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/CA+icZUV0F3r4G6pJ7Q=1nqgHb4SVEtdaA--9xWvdrxj_=e3...@mail.gmail.com
Bug#759809: linux-image-3.2.0-4-amd64: USB 3 devices fail, logging xHCI xhci_drop_endpoint called with disabled ep ffff8802165db0c0
On Tue, Sep 9, 2014 at 4:03 PM, Sedat Dilek sedat.di...@gmail.com wrote: Hi, this looks like what I reported in [wheezy][linux-3.2.60] Booting from a Debian system on an external USB-3.0 hdd hangs (see [1]). Unfortunately, there is no newer 3.2.y kernel for wheezy available [2]. Henrik, did you try a Linux-kernel = 3.2.61 (self-compiled)? Thanks. Regards, - Sedat - [1] https://lists.debian.org/debian-kernel/2014/08/msg00141.html [2] http://anonscm.debian.org/viewvc/kernel/dists/wheezy/linux/ The problem was solved by updating to linux-image-3.2.0-4-amd64 (3.2.63-2+deb7u1). - Sedat - -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/ca+iczuw2sxsmedkhzsfpk6w4epx5kmmi53dlttrpty_sexr...@mail.gmail.com
Bug#759809: linux-image-3.2.0-4-amd64: USB 3 devices fail, logging xHCI xhci_drop_endpoint called with disabled ep ffff8802165db0c0
On Sat, Nov 1, 2014 at 2:56 PM, Sedat Dilek sedat.di...@gmail.com wrote: [...] The problem was solved by updating to linux-image-3.2.0-4-amd64 (3.2.63-2+deb7u1). Just for the sake of completeness (and for Jari)... Here I have an ASM1042 chipset... [ lshw -sanitize ] ... *-pci:2 description: PCI bridge product: 6 Series/C200 Series Chipset Family PCI Express Root Port 5 vendor: Intel Corporation physical id: 1c.4 bus info: pci@:00:1c.4 version: b4 width: 32 bits clock: 33MHz capabilities: pci pciexpress msi pm normal_decode bus_master cap_list configuration: driver=pcieport resources: irq:16 memory:f050-f05f *-usb description: USB controller product: ASM1042 SuperSpeed USB Host Controller --- Look here! vendor: ASMedia Technology Inc. physical id: 0 bus info: pci@:03:00.0 version: 00 width: 64 bits clock: 33MHz capabilities: msi msix pm pciexpress xhci bus_master cap_list configuration: driver=xhci_hcd latency=0 resources: irq:16 memory:f050-f0507fff ... If you need more details, please let me know. - Sedat - -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/ca+iczuwvqt+dq6skpdsqmuwkw6hlbq9i5xcomdju6gnwfa0...@mail.gmail.com
Bug#759809: linux-image-3.2.0-4-amd64: USB 3 devices fail, logging xHCI xhci_drop_endpoint called with disabled ep ffff8802165db0c0
Hi, this looks like what I reported in [wheezy][linux-3.2.60] Booting from a Debian system on an external USB-3.0 hdd hangs (see [1]). Unfortunately, there is no newer 3.2.y kernel for wheezy available [2]. Henrik, did you try a Linux-kernel = 3.2.61 (self-compiled)? Thanks. Regards, - Sedat - [1] https://lists.debian.org/debian-kernel/2014/08/msg00141.html [2] http://anonscm.debian.org/viewvc/kernel/dists/wheezy/linux/ -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/CA+icZUVOGkzF=zrgasopse2dun1ugiwro0wcjzaawcxhtl_...@mail.gmail.com
[wheezy][linux-3.2.60] Booting from a Debian system on an external USB-3.0 hdd hangs
[ Please CC me - I am not subscribed to this ML ] Hi, I cannot boot with my current linux-3.2.y Debian/wheezy kernel as my system hangs and reports xhci problems. ( Unfortunately, nothing in the logs and no picture from my screen for you ). FYI: My system and bootloader etc. is on an external USB-3.0 (xhci) hdd. linux-image (3.2.57-3+deb7u1) was OK and (3.2.60-1+deb7u1) was 1st causing the problem. AFAICS this is the missing fix which is included in linux-3.2.61 [1]. commit 5dc2808c4729bf080487e61b80ee04e0fdb12a37 upstream. xhci: delete endpoints from bandwidth list before freeing whole device I have seen linux-3.2.62 was released recently. Can you integrate above fix as a hotfix or do you plan to push a linux-3.2.62 kernel soon? Thanks in advance. Regards, - Sedat - P.S.: If you need additional informations, please let me know. [1] http://git.kernel.org/cgit/linux/kernel/git/bwh/linux-3.2.y-queue.git /tree/releases/3.2.61/xhci-delete-endpoints-from-bandwidth-list-befor e-freeing-whole.patch -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/ca+iczuvena6c8xucrkqdjzt4mpmuddskuasbexroonz4zo6...@mail.gmail.com
Bug#695182: [RFC] Reproducible OOM with just a few sleeps
On Wed, Jan 16, 2013 at 1:22 AM, paul.sz...@sydney.edu.au wrote: Dear Sedat, ... it really makes sense to switch to x86_64 (amd64) architecture when you have a modern computer. Switching makes even more sense when you have more than 4GiB RAM. You seem to say that one should switch to amd64 (if hardware allows), even with less than 4GB RAM (where 32-bit non-PAE HIGHMEM4G kernel would work fine), and that one should definitely switch with over 4GB RAM. There would be no need or use for PAE kernels, which should be dropped. I think I agree. [ OK, you took the thread from LKML to the Debian bug, anyway ] Where I see problems is the fact that you are more or less forced to switch to 64-bit. Why? (Read my thoughts below.) The bigger problem I am seeing is that as most developers decided to go the 64-bit way the 32-bit path is no more tested correctly. When I insisted to run a 32-bit system I fell over so much UNTESTED software. I talked with a lot of developers around the Linux kernel and Debian world and those guys - if you ask them in private - would drop 32-bit entirely. To be honest - I am speaking of the x86 world and remember also darkly issues in early MULTIAAARGH support on Debian. ( As an example: Building a gcc upstream release tarball (unpatched!) in an multiarch environment. Look for my bug-reports if you are interested in. ) If there exist no more 32-bit x86 veterans... The world will turn around - approximately the same way and speed :-). So if you want to concentrate on working, make your decisions carefully! ( Noone pays you for fixing all the time your working OS - saying that as a longterm Debian/sid user. ) Just from my experiences. Regards, - Sedat - Thanks, Paul Paul Szabo p...@maths.usyd.edu.au http://www.maths.usyd.edu.au/u/psz/ School of Mathematics and Statistics University of SydneyAustralia --- Quoting in full for the benefit of 695...@bugs.debian.org : From sedat.di...@gmail.com Tue Jan 15 21:26:14 2013 Date: Tue, 15 Jan 2013 11:25:41 +0100 Subject: Re: [RFC] Reproducible OOM with just a few sleeps From: Sedat Dilek sedat.di...@gmail.com To: paul.sz...@sydney.edu.au, Paul Szabo p...@maths.usyd.edu.au Cc: LKML linux-ker...@vger.kernel.org, linux-mm linux...@kvack.org, Ben Hutchings b...@decadent.org.uk Hi Paul, I followed a bit the thread you started in [1]. As you might know i386 got eliminated in Linux-3.8. I had several discussions with the Debian kernel-team about the iN86 (N=4..6) and PAE kernel-flavours. On the one hand I can understand the reduction of linux-images especially for iN86. Even i486 is a bit unfirm as there is no much hardware around, but Debian will keep i486 for a while (release maintenance). Topic PAE: Unfortunately, I had a notebook with a Intel Centrino Banias CPU (no PAE) which should use the -486 kernel-flavour due to the Debian kernel-team. I played with some different kernel-setup which did not give me more benefit (openssl benchmarks etc.) The -686-pae kernel did run on my hardware, but as known with all the SMP-NO-OPs. Depending on the hardware, it really makes sense to switch to x86_64 (amd64) architecture when you have a modern computer. Switching makes even more sense when you have more than 4GiB RAM. IMHO using a -686-amd64 Debian kernel makes ZERO sense, real 64-Bit or die! I switched to 64-bit... and I switched from Debian/sid to Ubuntu/precise as well :-). ( NOTE: I am working here since April 2012 in a WUBI environment (no native Ubuntu Linux) :-). ) And I am building my kernels by myself. So I know very well whom to blame :-). Some last words: I had several fruitful or fruitless discussions with the Debian kernel-team, but I can confirm (with all my heart) this team makes a fantastic job. I can recommend you Ben's blog (recently I read a series about news in the Debian/wheezy kernel) if your world is Debian or Ubuntu (Debian != Ubuntu). Just my 0.02EUR (no British pound, here as well: when you are a member of the EU chose EUR not pound!). Regards, - Sedat - [1] http://marc.info/?t=13579617221r=1w=2 -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/ca+iczuwmxsrkpqklocjfquxbd0kfv7spzewvb7pvhq6epwg...@mail.gmail.com
Re: Mini-Howto: Build linux-kbuild-3.5
On Mon, Aug 20, 2012 at 2:03 AM, Ben Hutchings b...@decadent.org.uk wrote: On Fri, 2012-08-17 at 23:22 +0200, Sedat Dilek wrote: Hi, today, I was discussing with Michael and at some point he was complaining about the missing linux-kbuild-3.5 package in the Debian software archives. He was not able to install linux-headers-3.5 Debian package as it depends on it. So, this is for you, Michael. Hope this helps others as well. Thanks for pointing out the problems you had to work around, Sedat. I've fixed them in svn (I think) and will upload to experimental shortly. Hi Ben, A big thank you for taking care! Some notes and questions from here and IRC: [1] Dropped perf-UI support A wise decision for know. Yes, I have libgtk2.0-dev installed on my Debian/testing box. ( This mammoth should be a Build-Depends when enabling perf-UI (Have fun!). ) [2] [PATCH] perf: Fix include order for bison/flex-generated C files (sent to upstream) Thanks for the patch (see [1]). Is the misbehaviour with /usr/include/perf.h provided by firebird-dev package gone with that patch? Should firebird-dev rename its perf.h to something like firebird_perf.h to avoid such a misbehaviour? [3] Repacking Linux vanilla source now obsolete Thanks, this was really appreciated (see [2])! [ debian/bin/genorig.py (r19335) ] [CODE] Line #105: match = re.match(r'(^|.*/)(?Pdirlinux-\d+\.\d+(\.\d+)?(-\S+)?)\.tar(\.(?Pextension(bz2|gz)))?$', input_tar) [CODE] It is possible to get xz-tarball support (currently bzip2 or gzip)? Or to go a step further, what about an xz-compressed linux-tools orig-tarball? Here from my generated tarball (sizes in KiB): $ tar -xf linux-tools_3.5.orig.tar.gz $ tar -c --xz -f linux-tools_3.5.orig.tar.xz linux-tools-3.5/ $ du -k *.orig* 9280linux-tools_3.5.orig.tar.gz 6316linux-tools_3.5.orig.tar.xz Approx. 3MiB should be a good reason :-)? Note: The default compression-level of xz is -6 and -9 is maximum (see man xz). [4] Debian-Kernel Wiki There is a different handling of linux-2.6/linux-kbuild-2.6 VS. linux/linux-tools. Hector (zumbi) pointed to this already in [3]. Dunno, might be worth to write a separate Email on this topic? ( Just fell over this when people asked me for integration of my Mini-Howto, see [4]. ) Kind Regards, - Sedat - References: [1] http://marc.info/?l=linux-kernelm=134542880500616w=2 [2] http://anonscm.debian.org/viewvc/kernel/dists/trunk/linux-tools/debian/bin/genorig.py?r1=19335r2=19334 [3] http://wiki.debian.org/HowToRebuildAnOfficialDebianKernelPackage [4] http://wiki.debian.org/HowToRebuildAnOfficialDebianKernelPackage#The_story_of_linux-kbuild-2.6 Ben. -- Ben Hutchings The most exhausting thing in life is being insincere. - Anne Morrow Lindberg -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/ca+iczuudtr3mbzk4iabc09tw38o2f-p+0eyisz2pccvhnht...@mail.gmail.com
[PATCH] Support to extract upstream xz-compressed tarballs
Signed-off-by: Sedat Dilek sedat.di...@gmail.com --- debian/bin/genorig.py |2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/debian/bin/genorig.py b/debian/bin/genorig.py index 30a2e75..856d5eb 100755 --- a/debian/bin/genorig.py +++ b/debian/bin/genorig.py @@ -102,7 +102,7 @@ class Main(object): def upstream_extract(self, input_tar): self.log(Extracting tarball %s\n % input_tar) -match = re.match(r'(^|.*/)(?Pdirlinux-\d+\.\d+(\.\d+)?(-\S+)?)\.tar(\.(?Pextension(bz2|gz)))?$', input_tar) +match = re.match(r'(^|.*/)(?Pdirlinux-\d+\.\d+(\.\d+)?(-\S+)?)\.tar(\.(?Pextension(bz2|gz|xz)))?$', input_tar) if not match: raise RuntimeError(Can't identify name of tarball) -- 1.7.10.4 -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/1345469772-5435-1-git-send-email-sedat.di...@gmail.com
[PATCH] linux-tools: Add xz support
This patch supports xz-compressed upstream (vanilla) tarball source
to generate an xz-compressed linux-tools orig-tarball.
Furthermore, patches are now allowed in xz-format, too.
As suggested by Ben I have replaced all occurences of tar with -a
option for compressing/decompressing tarballs.
From man tar(1):
-a, --auto-compress
use archive suffix to determine the compression program
After applying this patch, follow these instructions (here: Linux-v3.5):
$ cd linux-tools/
$ make -f debian/rules debian/control-real
$ ./debian/bin/genorig.py /path/to/linux-3.5.tar.xz
$ fakeroot debian/rules orig
$ debuild
Tested against linux-tools SVN trunk revision 19348.
Signed-off-by: Sedat Dilek sedat.di...@gmail.com
---
debian/bin/genorig.py | 18 --
debian/rules |4 ++--
2 files changed, 10 insertions(+), 12 deletions(-)
diff --git a/debian/bin/genorig.py b/debian/bin/genorig.py
index 30a2e75..2465209 100755
--- a/debian/bin/genorig.py
+++ b/debian/bin/genorig.py
@@ -66,7 +66,7 @@ class Main(object):
self.log('Using source name %s, version %s, dfsg %s\n' % (source,
version.upstream, self.version_dfsg))
self.orig = '%s-%s' % (source, version.upstream)
-self.orig_tar = '%s_%s.orig.tar.gz' % (source, version.upstream)
+self.orig_tar = '%s_%s.orig.tar.xz' % (source, version.upstream)
self.tag = 'v' + re.sub(r^(\d+\.\d+)\.0, r\1,
version.upstream.replace('~', '-'))
@@ -92,7 +92,7 @@ class Main(object):
'--prefix=temp/', self.tag],
cwd=input_repo,
stdout=subprocess.PIPE)
-extract_proc = subprocess.Popen(['tar', '-xf', '-'], cwd=self.dir,
+extract_proc = subprocess.Popen(['tar', '-xaf', '-'], cwd=self.dir,
stdin=archive_proc.stdout)
ret1 = archive_proc.wait()
@@ -102,15 +102,11 @@ class Main(object):
def upstream_extract(self, input_tar):
self.log(Extracting tarball %s\n % input_tar)
-match =
re.match(r'(^|.*/)(?Pdirlinux-\d+\.\d+(\.\d+)?(-\S+)?)\.tar(\.(?Pextension(bz2|gz)))?$',
input_tar)
+match =
re.match(r'(^|.*/)(?Pdirlinux-\d+\.\d+(\.\d+)?(-\S+)?)\.tar(\.(?Pextension(bz2|gz|xz)))?$',
input_tar)
if not match:
raise RuntimeError(Can't identify name of tarball)
-cmdline = ['tar', '-xf', input_tar, '-C', self.dir]
-if match.group('extension') == 'bz2':
-cmdline.append('-j')
-elif match.group('extension') == 'gz':
-cmdline.append('-z')
+cmdline = ['tar', '-xaf', input_tar, '-C', self.dir]
if subprocess.Popen(cmdline).wait():
raise RuntimeError(Can't extract tarball)
@@ -119,7 +115,7 @@ class Main(object):
def upstream_patch(self, input_patch):
self.log(Patching source with %s\n % input_patch)
-match =
re.match(r'(^|.*/)patch-\d+\.\d+\.\d+(-\S+?)?(\.(?Pextension(bz2|gz)))?$',
input_patch)
+match =
re.match(r'(^|.*/)patch-\d+\.\d+\.\d+(-\S+?)?(\.(?Pextension(bz2|gz|xz)))?$',
input_patch)
if not match:
raise RuntimeError(Can't identify name of patch)
cmdline = []
@@ -127,6 +123,8 @@ class Main(object):
cmdline.append('bzcat')
elif match.group('extension') == 'gz':
cmdline.append('zcat')
+elif match.group('extension') == 'xz':
+cmdline.append('xzcat')
else:
cmdline.append('cat')
cmdline.append(input_patch)
@@ -177,7 +175,7 @@ class Main(object):
except OSError:
pass
self.log(Generate tarball %s\n % out)
-cmdline = ['tar -czf', out, '-C', self.dir, self.orig]
+cmdline = ['tar -caf', out, '-C', self.dir, self.orig]
try:
if os.spawnv(os.P_WAIT, '/bin/sh', ['sh', '-c', '
'.join(cmdline)]):
raise RuntimeError(Can't patch source)
diff --git a/debian/rules b/debian/rules
index 28304f7..3a21ac5 100755
--- a/debian/rules
+++ b/debian/rules
@@ -20,7 +20,7 @@ $(STAMPS_DIR):
@[ -d $@ ] || mkdir $@
DIR_ORIG = ../orig/$(SOURCE)-$(VERSION)
-TAR_ORIG_NAME = $(SOURCE)_$(VERSION).orig.tar.gz
+TAR_ORIG_NAME = $(SOURCE)_$(VERSION).orig.tar.xz
TAR_ORIG = $(firstword $(wildcard ../$(TAR_ORIG_NAME)) $(wildcard
../orig/$(TAR_ORIG_NAME)))
orig: $(DIR_ORIG)
@@ -31,7 +31,7 @@ ifeq ($(TAR_ORIG),)
$(error Cannot find orig tarball $(TAR_ORIG_NAME))
else
mkdir -p ../orig
- tar -C ../orig -xzf $(TAR_ORIG)
+ tar -C ../orig -xaf $(TAR_ORIG)
endif
maintainerclean:
--
1.7.10.4
--
To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive:
http://lists.debian.org/1345480843-24447-1-git-send-email-sedat.di
Re: Mini-Howto: Build linux-kbuild-3.5
On Sat, Aug 18, 2012 at 5:58 PM, Sedat Dilek sedat.di...@gmail.com wrote: [ Please CC me I am not subscribed to debian-kernel ML ] [ QUOTE ] [...] Thanks for writing this up. Maybe you would want to put this up on wiki. http://wiki.debian.org/HowToRebuildAnOfficialDebianKernelPackage [ /QUOTE ] [...] An idea could be to attach my Mini-Howto TXT-file (v2 attached)... Added a reference to my Mini-Howto (see [1]). - Sedat - [1] http://wiki.debian.org/HowToRebuildAnOfficialDebianKernelPackage#The_story_of_linux-kbuild-2.6 -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/CA+icZUWyB-syQ_A0PGRApJkmBDZZbaksjMz2=gmelzwvtpu...@mail.gmail.com
Re: Mini-Howto: Build linux-kbuild-3.5
[ Please CC me I am not subscribed to debian-kernel ML ]
[ QUOTE ]
On Saturday 18 August 2012 02:52 AM, Sedat Dilek wrote:
Hi,
today, I was discussing with Michael and at some point he was
complaining about the missing linux-kbuild-3.5 package in the
Debian software archives. He was not able to install
linux-headers-3.5 Debian package as it depends on it.
So, this is for you, Michael.
Hope this helps others as well.
Thanks for writing this up. Maybe you would want to put this up on wiki.
http://wiki.debian.org/HowToRebuildAnOfficialDebianKernelPackage
[ /QUOTE ]
Please, check the history of contributions (Info button) :-).
Then you will see I have done a lot of in the section The story of
linux-kbuild-2.6.
The appropriate wiki is always a good place to put such a step-by-step
build-instructions.
BUT I am describing a corner-case:
The move to linux-tools affects Debian-kernels =3.2, so you have to
split within this wiki-section as there is for example Debian/lenny
with 2.6.32-kernel.
And... my base is a Linux vanilla source - I never liked that
DFSG-pruned tarball - it's additional work.
I respect the Debian philosophy - but I have the work - it's my time.
The GPL-v2 Linux vanilla kernel-source is easier from handling for me
(as I have it stored on my local hdd).
An idea could be to attach my Mini-Howto TXT-file (v2 attached)...
- Sedat (dileks on IRC) -
# Filename: Mini-Howto_Build_linux-kbuild-3-5.txt
# Author: Sedat Dilek sedat.di...@gmail.com
# Created: 17-Aug-2012
# Last Changed: 18-Aug-2012 (v2)
# NOTE-1: Source for linux-kbuild-3.5 Debian binaries is linux-tools!
# NOTE-2: linux-headers-3.5 package depends on linux-kbuild-3.5 (1st
exists in the archives 2nd not)
# ( Known problem: Same procedure as with every linux-trunk
release, Mrs. Sophie? )
# WARNING: firebird-dev breaks perf compilation due to sh*t shipped within
/usr/include/perf.h !
# XXX: Workaround: Deinstall firebird-dev package!
# EXPORTS
export LANG=C
export LC_ALL=C
# SOME USEFUL VARIABLES
### Package names
PKG_NAME=linux-tools
PKG_NAME_LATEST=${PKG_NAME}-latest
### Version strings
PKG_VER=3.5
KERNEL_VERSION=$PKG_VER
PKG_VER_LATEST=3.4-1~experimental.1
### Tarballs
KERNEL_TARBALL=linux-${KERNEL_VERSION}.tar.xz
DEBIAN_TARBALL=${PKG_NAME}_${PKG_VER}.orig.tar.gz
### XXX: Workaround for new naming-scheme from upstream (tarball: 2 digits VS.
Makefile: 3 digits)
KERNEL_VERSION_REPACKED=${PKG_VER}.0
KERNEL_TARBALL_REPACKED=linux-${KERNEL_VERSION_REPACKED}.tar.gz
### URLs
KERNEL_FTP_URL=http://ftp.kernel.org/pub/linux/kernel/v3.x;
DEBIAN_SVN_URL=svn://svn.debian.org/kernel/releases/${PKG_NAME}
### Working and build directories
WORK_DIR=${HOME}/src/${PKG_NAME}
BUILD_DIR=${WORK_DIR}/${PKG_NAME}-${PKG_VER}
# PREPARE DEBIAN PACKAGE
### Create and enter working directory
[ -d $WORK_DIR ] || mkdir -p $WORK_DIR
cd $WORK_DIR
### Download Linux vanilla source file
wget $KERNEL_FTP_URL/$KERNEL_TARBALL
### Checkout $PKG_NAME_LATEST from $DEBIAN_SVN_URL
svn co $DEBIAN_SVN_URL/$PKG_VER_LATEST $PKG_NAME_LATEST
### Enter the arena
cd $PKG_NAME_LATEST
### Remove unwanted SVN-repository related files and dirs
rm -v -rf $(find ./ -name .svn)
### Generate missing debian/control file to satisfy dch (included in devscripts
package)
make -f debian/rules debian/control-real
### Bump package-name and debian-version
dch -i
### Re-create and check debian/control file for linux-kbuild (should contain
$PKG_VER)
make -f debian/rules debian/control-real
grep linux-kbuild debian/control
### XXX: Workaround: Repack Linux vanilla tarball
cd $WORK_DIR
tar -xf $KERNEL_TARBALL
mv linux-${KERNEL_VERSION} linux-${KERNEL_VERSION_REPACKED}
tar -c --gzip -f $KERNEL_TARBALL_REPACKED linux-${KERNEL_VERSION_REPACKED}
rm -rf linux-${KERNEL_VERSION_REPACKED}
### Generate DFSG-pruned $DEBIAN_TARBALL
cd $PKG_NAME_LATEST
./debian/bin/genorig.py ../$KERNEL_TARBALL_REPACKED
### Unpack $DEBIAN_TARBALL
cd $WORK_DIR
tar -xf orig/$DEBIAN_TARBALL
# Copy debian-dir from SVN checkout
cp -av ${PKG_NAME_LATEST}/debian $BUILD_DIR
### Delete symlink to orig/$DEBIAN_TARBALL, move the real file and delete
orig-dir
rm -v -f $DEBIAN_TARBALL
mv -v orig/$DEBIAN_TARBALL ./
rm -v -rf orig/
# BUILD DEBIAN PACKAGE
### Enter build-dir
cd $BUILD_DIR
### Check build-dependencies (run 'apt-get build-dep linux-tools' to satisfy
build-deps)
dpkg-checkbuilddeps
### Start the build (alternatively: run 'dpkg-buildpackage -us -uc')
debuild
# CLEAN-UP
cd $WORK_DIR
### Remove kernel tarballs
rm -v -f $KERNEL_TARBALL
rm -v -f $KERNEL_TARBALL_REPACKED
### Remove directory from SVN checkout
rm -v -rf $PKG_NAME_LATEST
### Remove build-dir
rm -v -rf $BUILD_DIR
# INSTALL DEBIAN PACKAGE
### Install linux-kbuild package manually
dpkg -i linux-kbuild-${PKG_VER}_*.deb
Mini-Howto: Build linux-kbuild-3.5
Hi,
today, I was discussing with Michael and at some point he was
complaining about the missing linux-kbuild-3.5 package in the Debian
software archives.
He was not able to install linux-headers-3.5 Debian package as it depends on it.
So, this is for you, Michael.
Hope this helps others as well.
Regards,
- Sedat -
#!/bin/sh
# Filename: Mini-Howto_Build_linux-kbuild-3-5.txt
# Author: Sedat Dilek sedat.di...@gmail.com
# Created: 17-Aug-2012
# Last Changed: 17-Aug-2012
# NOTE-1: Source for linux-kbuild-3.5 Debian binaries is linux-tools!
# NOTE-2: linux-headers-3.5 package depends on linux-kbuild-3.5 (1st
exists in the archives 2nd not)
# ( Known problem: Same procedure as with every linux-trunk
release, Mrs. Sophie? )
# WARNING: firebird-dev breaks perf compilation due to sh*t shipped within
/usr/include/perf.h !
# XXX: Workaround: Deinstall firebird-dev package!
# EXPORTS
export LANG=C
export LC_ALL=C
# SOME USEFUL VARIABLES
### Package names
PKG_NAME=linux-tools
PKG_NAME_LATEST=${PKG_NAME}-latest
### Version strings
PKG_VER=3.5
KERNEL_VERSION=$PKG_VER
KERNEL_VERSION_REPACKED=${PKG_VER}.0
PKG_VER_LATEST=3.4-1~experimental.1
### Tarballs
KERNEL_TARBALL=linux-${KERNEL_VERSION}.tar.xz
DEBIAN_TARBALL=${PKG_NAME}_${PKG_VER}.orig.tar.gz
# XXX: Workaround for new naming-scheme (tarball: 2 digits VS. Makefile: 3
digits)
KERNEL_TARBALL_REPACKED=linux-${KERNEL_VERSION_REPACKED}.tar.gz
### URLs
KERNEL_FTP_URL=http://ftp.kernel.org/pub/linux/kernel/v3.x;
DEBIAN_SVN_URL=svn://svn.debian.org/kernel/releases/${PKG_NAME}
### Working and build directorie
WORK_DIR=${HOME}/src/${PKG_NAME}
BUILD_DIR=${WORK_DIR}/${PKG_NAME}-${PKG_VER}
# PREPARE DEBIAN PACKAGE
### Create and enter working directory
[ -d $WORK_DIR ] || mkdir $WORK_DIR
cd $WORK_DIR
### Download Linux vanilla source file
wget $KERNEL_FTP_URL/$KERNEL_TARBALL
### Checkout $PKG_NAME_LATEST from $DEBIAN_SVN_URL
svn co $DEBIAN_SVN_URL/$PKG_VER_LATEST $PKG_NAME_LATEST
### Enter the arena
cd $PKG_NAME_LATEST
### Remove unwanted SVN-repository related files and dirs
rm -v -rf $(find ./ -name .svn)
### Generate missing debian/control file to satisfy dch (included in devscripts
package)
make -f debian/rules debian/control-real
### Bump package-name and debian-version
dch -i
### Re-create and check debian/control file for linux-kbuild (should contain
$PKG_VER)
make -f debian/rules debian/control-real
grep linux-kbuild debian/control
### XXX: Workaround: Repack Linux vanilla tarball
cd $WORK_DIR
tar -xf $KERNEL_TARBALL
mv linux-${KERNEL_VERSION} linux-${KERNEL_VERSION_REPACKED}
tar -c --gzip -f $KERNEL_TARBALL_REPACKED linux-${KERNEL_VERSION_REPACKED}
rm -rf linux-${KERNEL_VERSION_REPACKED}
### Generate DFSG-pruned $DEBIAN_TARBALL
cd $PKG_NAME_LATEST
./debian/bin/genorig.py ../$KERNEL_TARBALL_REPACKED
### Unpack $DEBIAN_TARBALL
cd $WORK_DIR
tar -xf orig/$DEBIAN_TARBALL
# Copy debian-dir from SVN checkout
cp -av ${PKG_NAME_LATEST}/debian $BUILD_DIR
### Delete symlink to orig/$DEBIAN_TARBALL, move the real file and delete
orig-dir
rm -v -f $DEBIAN_TARBALL
mv -v orig/$DEBIAN_TARBALL ./
rm -v -rf orig/
# BUILD DEBIAN PACKAGE
### Enter build-dir
cd $BUILD_DIR
### Check build-dependencies (run 'apt-get build-dep linux-tools' to satisfy
build-deps)
dpkg-checkbuilddeps
### Start the build (alternatively: run 'dpkg-buildpackage -us -uc')
debuild
# CLEAN-UP
cd $WORK_DIR
### Remove kernel tarballs
rm -v -f $KERNEL_TARBALL
rm -v -f $KERNEL_TARBALL_REPACKED
### Remove directory from SVN checkout
rm -v -rf $PKG_NAME_LATEST
### Remove build-dir
rm -v -rf $BUILD_DIR
# INSTALL DEBIAN PACKAGE
### Install linux-kbuild package manually
dpkg -i linux-kbuild-${PKG_VER}_*.deb
Re: [linux-2.6] 3.1.0-1: Cosmetic fix: [SCSI] Silencing 'killing requests for dead queue'
UPDATE: Fix will be in linux-3.1.5 as scsi-silencing-killing-requests-for-dead-queue.patch [1] http://git.kernel.org/?p=linux/kernel/git/stable/stable-queue.git;a=blob;f=queue-3.1/series#l52 On Fri, Nov 25, 2011 at 12:42 PM, Sedat Dilek sedat.di...@googlemail.com wrote: Hi, I had these scsi: killing requests for dead queue in linux-next for a while. It's fixed there and in mainline, now. Unfortunately, I saw these message when booting into linux-image-3.1.0-1-686-pae. The upstream fix is commit 745718132c ([SCSI] Silencing 'killing requests for dead queue'). For details please see [1]. If you like you can apply it. Regards, - Sedat - [1] http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git;a=commitdiff;h=745718132c3c7cac98a622b610e239dcd5217f71 -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/CA+icZUUGL7Sktr=B_F9O1S2=0k-329Uip4wJG3k=0kbp26m...@mail.gmail.com
[linux-2.6] 3.1.0-1: Cosmetic fix: [SCSI] Silencing 'killing requests for dead queue'
Hi, I had these scsi: killing requests for dead queue in linux-next for a while. It's fixed there and in mainline, now. Unfortunately, I saw these message when booting into linux-image-3.1.0-1-686-pae. The upstream fix is commit 745718132c ([SCSI] Silencing 'killing requests for dead queue'). For details please see [1]. If you like you can apply it. Regards, - Sedat - [1] http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git;a=commitdiff;h=745718132c3c7cac98a622b610e239dcd5217f71 -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/CA+icZUU=BCDDBKGT_ZtmrUmE6Lkw=JJLo3uZr4t+x=t1z4a...@mail.gmail.com
[PATCH] i386: Fix flavours in [686-pae_xen] define
Signed-off-by: Sedat Dilek sedat.di...@gmail.com --- debian/config/i386/defines |3 +-- 1 files changed, 1 insertions(+), 2 deletions(-) diff --git a/debian/config/i386/defines b/debian/config/i386/defines index 466949b..a2da916 100644 --- a/debian/config/i386/defines +++ b/debian/config/i386/defines @@ -43,8 +43,7 @@ enabled: false [686-pae_xen] dom0-support: true flavours: - i386 - amd64 + 686-pae [amd64_description] hardware: 64-bit PCs -- 1.7.6 -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/1311518116-11223-1-git-send-email-sedat.di...@gmail.com
Re: [debian-kernel] RT patchset v3.0-rt3 for i386 against vanilla Linux-3.0
As an addendum:
ARCH=i386
FEATURESET=rt
# Generate missing linux-headers-*-common package
fakeroot make -f debian/rules.gen binary-arch_${ARCH}_${FEATURESET}_real
- Sedat -
2011/7/24 Sedat Dilek sedat.di...@googlemail.com:
Hi,
I played a bit with the new Real-Time (RT) featureset on i386.
The RT featureset with v3.0-rt2 patch was enabled by default with
(3.0.0-1) for amd64 arch.
Attached is my patchset enabling RT for i386 and 686-pae flavour with
an updated v3.0-rt3 patch included.
Furthermore, I used gcc-4.6 as default-compiler (default in Debian/sid).
Also, I disabled build of the debug linux-image (explodes build-dir up
to ~6.3GiB).
My base is a Linux-3.0 vanilla tarball.
Most interesting patch should be 0004.
Have fun!
Regards,
- Sedat -
### INSTRUCTIONS
# Download and unpack Linux-3.0 tarball
$ wget http://www.kernel.org/pub/linux/kernel/v3.0/linux-3.0.tar.bz2
$ tar -xf linux-3.0.tar.bz2
# Checkout debian directory of 3.0.0-1 release from SVN repository
$ cd linux-3.0
$ svn co svn://svn.debian.org/svn/kernel/releases/linux-2.6/3.0.0-1/debian
# Apply rt3-patchset
$ patch -Np1 -i /path/to/patchset/$patch-1
( ...$patch-2 ...$patch-3 etc. )
# Create missing control and rules.gen files.
fakeroot make -f debian/rules debian/control-real
ARCH=i386
FEATURESET=rt
FLAVOUR=686-pae
SINGLE_KERNEL_VARIANT=${ARCH}_${FEATURESET}_${FLAVOUR}
# Check setup of 'i386_rt_686-pae' single kernel variant
$ fakeroot make -f debian/rules.gen setup_$SINGLE_KERNEL_VARIANT
# Start build of 'i386_rt_686-pae' single kernel variant
$ fakeroot make -f debian/rules.gen binary-arch_$SINGLE_KERNEL_VARIANT
### STATISTICS
Build-time: approx. 02:15h on IBM ThinkPad T41p (Intel Pentium-M, 1GiB
RAM, 60GiB HDD, etc.)
Disc-Usage: approx. 1.5GiB
- Sedat -
--
To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive:
http://lists.debian.org/CA+icZUUPbipMbTgnM9C28APst=E4rq6=ofywbuxd9hcooyt...@mail.gmail.com
Re: [PATCH] i386: Fix flavours in [686-pae_xen] define
On Sun, Jul 24, 2011 at 6:41 PM, Ben Hutchings b...@decadent.org.uk wrote: On Sun, 2011-07-24 at 16:35 +0200, Sedat Dilek wrote: Signed-off-by: Sedat Dilek sedat.di...@gmail.com --- debian/config/i386/defines | 3 +-- 1 files changed, 1 insertions(+), 2 deletions(-) diff --git a/debian/config/i386/defines b/debian/config/i386/defines index 466949b..a2da916 100644 --- a/debian/config/i386/defines +++ b/debian/config/i386/defines @@ -43,8 +43,7 @@ enabled: false [686-pae_xen] dom0-support: true flavours: - i386 - amd64 + 686-pae This section defines the flavours of the _hypervisor_ that can be used with this flavour of the kernel. The current value is correct. Ben. -- Ben Hutchings I'm always amazed by the number of people who take up solipsism because they heard someone else explain it. - E*Borg on alt.fan.pratchett A bit confusing as i386 is not a defined flavour or is this a different flavours variable than 486, 686-pae or amd64? - Sedat - -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/ca+iczuwjn5tzcfld-b02zebh_k-g2+q0yft576_feeowbzb...@mail.gmail.com
Bug#624372: linux-source-2.6.39: Fix warnings for invalid symbol values (now boolean)
Package: linux-source-2.6.39 Version: 2.6.39~rc4-1~experimental.1 Severity: normal Tags: patch Hi, attached patch against debian-dir from linux-2.6/trunk (r17251) fixes these warnings: .config:555:warning: symbol value 'm' invalid for LEDS_CLASS .config:917:warning: symbol value 'm' invalid for MFD_WM8994 .config:2292:warning: symbol value 'm' invalid for BT_L2CAP .config:2293:warning: symbol value 'm' invalid for BT_SCO - Sedat - -- System Information: Debian Release: wheezy/sid APT prefers unstable APT policy: (500, 'unstable'), (200, 'experimental') Architecture: i386 (i686) Kernel: Linux 2.6.39-rc4-686-pae Locale: LANG=de_DE.UTF-8, LC_CTYPE=de_DE.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Fix-warnings-for-invalid-symbol-values-now-boolean.patch Description: plain/text
build-system: Interaction changelog - abiname?
Hi, I am using a slightly modified version of the kernel-buildsystem from trunk which you offer to build my linux-next kernels. Unfortunately, even I use UNRELEASED as Distribution my $ABINAME is not accepted, I have these setttings: $ head -1 debian/changelog linux-2.6 (2.6.39~rc1-1~next20110331.dileks1) UNRELEASED; urgency=low $ head -2 debian/config/defines [abi] abiname: next20110331-1 If I remove ~rc1 means (2.6.39-1~next20110331.dileks.1) from debian/changelog, the desired ABINAME=next20110331-1 successfully is set. Using ~rc1 results in identical package names whereas the debian-version differs, means uname -r gives the same output. This is uncool when you have different patch-series or kernel-configs for testing and results in overwriting old package (or boot a totally different kernel, remove and install desired package). I have looked and try to understand the interaction between changelog - abiname. For example into debian/bin/gencontrol.py and debian/lib/python/debian_linux/debian.py. I would lie if I have understood the code. Can you give some help? Regards, - Sedat - -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/aanlktikjg55ozhib7gsfg2hgygduqf5dtbakfcbhe...@mail.gmail.com
Re: build-system: Interaction changelog - abiname?
[ Please CC me I am not subscribed to the list ]
[QUOTE]
On Thu, Mar 31, 2011 at 12:04:13PM +0200, Sedat Dilek wrote:
linux-2.6 (2.6.39~rc1-1~next20110331.dileks1) UNRELEASED; urgency=low
This version splits into 2.6.39~rc1 and 1~next20110331.dileks1.
$ head -2 debian/config/defines
[abi]
abiname: next20110331-1
If I remove ~rc1 means (2.6.39-1~next20110331.dileks.1) from
debian/changelog, the desired ABINAME=next20110331-1 successfully is
set.
A version of the A.B.C~rcX always forces the abiname. There is no
reason for Debian kernel to do it different.
Can you give some help?
You want a different version. Something like 2.6.39~rc1+next20110331-1.
Or remove the check in your tree.
Bastian
[/QUOTE]
I tried with:
$ head -1 debian/changelog
linux-2.6 (2.6.39~rc1+next20110331-1) UNRELEASED; urgency=low
$ head -2 debian/config/defines
[abi]
abiname: next20110331-1
NOPE.
$ grep -i ABINAME= setup.log
make -f debian/rules.real setup-flavour ABINAME='' ARCH='i386'
COMPILER='gcc-4.4' FEATURESET='none' FLAVOUR='686'
INITRD_CMD='update-initramfs' KCONFIG='debian/config/config
debian/config/kernelarch-x86/config
debian/config/kernelarch-x86/config-arch-32
debian/config/i386/none/config.686' KERNEL_ARCH='x86'
LOCALVERSION='-686' LOCALVERSION_HEADERS='' LOCALVERSION_IMAGE='-686'
MAJOR='2.6' MODULES='True' SOURCEVERSION='2.6.39~rc1+next20110331-1'
TYPE='plain' UPSTREAMVERSION='2.6.39-rc1+next20110331'
VERSION='2.6.39'
The next what I see with your version-string, what is the name of the
resulting orig-tarball?
The idea behind mine was X~rc1.orig.tar.{gz|bz|xz}.
What do you mean by Or remove the check in your tree.?
Where exactly?
- Sedat -
--
To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive:
http://lists.debian.org/AANLkTin7uVDe=cypnqhdosynmoyoszp9n0y2tp_c2...@mail.gmail.com
Re: build-system: Interaction changelog - abiname?
On Thu, Mar 31, 2011 at 2:56 PM, Sedat Dilek sedat.di...@googlemail.com wrote:
[ Please CC me I am not subscribed to the list ]
[QUOTE]
On Thu, Mar 31, 2011 at 12:04:13PM +0200, Sedat Dilek wrote:
linux-2.6 (2.6.39~rc1-1~next20110331.dileks1) UNRELEASED; urgency=low
This version splits into 2.6.39~rc1 and 1~next20110331.dileks1.
$ head -2 debian/config/defines
[abi]
abiname: next20110331-1
If I remove ~rc1 means (2.6.39-1~next20110331.dileks.1) from
debian/changelog, the desired ABINAME=next20110331-1 successfully is
set.
A version of the A.B.C~rcX always forces the abiname. There is no
reason for Debian kernel to do it different.
Can you give some help?
You want a different version. Something like 2.6.39~rc1+next20110331-1.
Or remove the check in your tree.
Bastian
[/QUOTE]
I tried with:
$ head -1 debian/changelog
linux-2.6 (2.6.39~rc1+next20110331-1) UNRELEASED; urgency=low
$ head -2 debian/config/defines
[abi]
abiname: next20110331-1
NOPE.
$ grep -i ABINAME= setup.log
make -f debian/rules.real setup-flavour ABINAME='' ARCH='i386'
COMPILER='gcc-4.4' FEATURESET='none' FLAVOUR='686'
INITRD_CMD='update-initramfs' KCONFIG='debian/config/config
debian/config/kernelarch-x86/config
debian/config/kernelarch-x86/config-arch-32
debian/config/i386/none/config.686' KERNEL_ARCH='x86'
LOCALVERSION='-686' LOCALVERSION_HEADERS='' LOCALVERSION_IMAGE='-686'
MAJOR='2.6' MODULES='True' SOURCEVERSION='2.6.39~rc1+next20110331-1'
TYPE='plain' UPSTREAMVERSION='2.6.39-rc1+next20110331'
VERSION='2.6.39'
The next what I see with your version-string, what is the name of the
resulting orig-tarball?
The idea behind mine was X~rc1.orig.tar.{gz|bz|xz}.
What do you mean by Or remove the check in your tree.?
Where exactly?
- Sedat -
With this Workaround ABINAME is correct set.
- Sedat -
diff --git a/debian/bin/gencontrol.py b/debian/bin/gencontrol.py
index a0cdfec..422a88f 100755
--- a/debian/bin/gencontrol.py
+++ b/debian/bin/gencontrol.py
@@ -323,10 +323,11 @@ class Gencontrol(Base):
versions.append(i.version)
self.versions = versions
version = self.version = self.changelog[0].version
-if self.version.linux_modifier is not None:
-self.abiname = ''
-else:
-self.abiname = '-%s' % self.config['abi',]['abiname']
+ # XXX: Workaround, remove this check
+ #if self.version.linux_modifier is not None:
+ #self.abiname = ''
+ #else:
+ self.abiname = '-%s' % self.config['abi',]['abiname']
self.vars = {
'upstreamversion': self.version.linux_upstream,
'version': self.version.linux_version,
--
To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive:
http://lists.debian.org/AANLkTimnYFU7jL4rkeX80wTH70=FgmmfOTE4x=gwd...@mail.gmail.com
[linux-2.6] trunk-r16272: Fix up some configs and speakup patch broken for 2.6.36-rc4
Hi,
I already pointed out on IRC that some CONFIG_* has to be updated for
2.6.36-rc4.
See my attached patches.
As it might be important: My linux-2.6/debian directory is from trunk-r16272.
Unfortunately, the shipped speakup patch(es) are somehow broken:
s...@tbox:~/src/linux-2.6$ grep -i speakup BROKEN_speakup_build.log
LD drivers/staging/speakup/built-in.o
CC [M] drivers/staging/speakup/speakup_acntsa.o
CC [M] drivers/staging/speakup/speakup_acntpc.o
CC [M] drivers/staging/speakup/speakup_apollo.o
CC [M] drivers/staging/speakup/speakup_audptr.o
CC [M] drivers/staging/speakup/speakup_bns.o
CC [M] drivers/staging/speakup/speakup_dectlk.o
CC [M] drivers/staging/speakup/speakup_decext.o
CC [M] drivers/staging/speakup/speakup_dtlk.o
CC [M] drivers/staging/speakup/speakup_keypc.o
CC [M] drivers/staging/speakup/speakup_ltlk.o
CC [M] drivers/staging/speakup/speakup_soft.o
CC [M] drivers/staging/speakup/speakup_spkout.o
CC [M] drivers/staging/speakup/speakup_txprt.o
CC [M] drivers/staging/speakup/speakup_dummy.o
CC [M] drivers/staging/speakup/buffers.o
/home/sd/src/linux-2.6/linux-2.6.36-rc4/debian/build/source_i386_none/drivers/staging/speakup/buffers.c:
In function ‘speakup_start_ttys’:
/home/sd/src/linux-2.6/linux-2.6.36-rc4/debian/build/source_i386_none/drivers/staging/speakup/buffers.c:33:
error: ‘struct vc_data’ has no member named ‘vc_tty’
/home/sd/src/linux-2.6/linux-2.6.36-rc4/debian/build/source_i386_none/drivers/staging/speakup/buffers.c:34:
error: ‘struct vc_data’ has no member named ‘vc_tty’
/home/sd/src/linux-2.6/linux-2.6.36-rc4/debian/build/source_i386_none/drivers/staging/speakup/buffers.c:
In function ‘speakup_stop_ttys’:
/home/sd/src/linux-2.6/linux-2.6.36-rc4/debian/build/source_i386_none/drivers/staging/speakup/buffers.c:56:
error: ‘struct vc_data’ has no member named ‘vc_tty’
/home/sd/src/linux-2.6/linux-2.6.36-rc4/debian/build/source_i386_none/drivers/staging/speakup/buffers.c:57:
error: ‘struct vc_data’ has no member named ‘vc_tty’
make[7]: *** [drivers/staging/speakup/buffers.o] Fehler 1
make[6]: *** [drivers/staging/speakup] Fehler 2
Some days back, I looked into speakup GIT repository, so the extracted
patch seems to be correct on first sight but I guess no more fitting
to 2.6.36-rc4.
Might be it's fixed in the meantime in upstream GIT.
As a temporary workaround (using i386_none_686 for builds) I have
disabled speakup in:
[ debian/config/i386/none/config.686 ]
...
# CONFIG_DEBUG_HIGHMEM is not set
##
## file: drivers/staging/speakup/Kconfig
##
# CONFIG_SPEAKUP is not set
-EOF-
Hope I could help and you have time for a closer look into my reported issues.
Kind Regards,
- Sedat (aka dile{ks,X} on IRC) -
P.S.: Yes, I know 2.6.36-rc4 support is WIP (UNRELEASED).
From 244692cc6f6fad05d1058bf9ca123535d8014368 Mon Sep 17 00:00:00 2001
From: Sedat Dilek sedat.di...@gmail.com
Date: Mon, 13 Sep 2010 05:46:38 +0200
Subject: [PATCH] Replace CONFIG_INOTIFY by CONFIG_FANOTIFY
---
debian/config/config |6 +-
1 files changed, 5 insertions(+), 1 deletions(-)
diff --git a/debian/config/config b/debian/config/config
index 28336b2..398865a 100644
--- a/debian/config/config
+++ b/debian/config/config
@@ -3175,10 +3175,14 @@ CONFIG_DNOTIFY=y
##
## file: fs/notify/inotify/Kconfig
##
-CONFIG_INOTIFY=y
CONFIG_INOTIFY_USER=y
##
+## file: fs/notify/fanotify/Kconfig
+##
+CONFIG_FANOTIFY=y
+
+##
## file: fs/ntfs/Kconfig
##
CONFIG_NTFS_FS=m
--
1.7.2.3
From 2c329cbf3c17b46f6de91f2012b96489d91facfa Mon Sep 17 00:00:00 2001
From: Sedat Dilek sedat.di...@gmail.com
Date: Mon, 13 Sep 2010 05:48:27 +0200
Subject: [PATCH] Unset CONFIG_MFD_WM8994
---
debian/config/config |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/debian/config/config b/debian/config/config
index 398865a..1de3909 100644
--- a/debian/config/config
+++ b/debian/config/config
@@ -1336,7 +1336,7 @@ CONFIG_MFD_SM501=m
CONFIG_HTC_PASIC3=m
CONFIG_TPS65010=m
CONFIG_MFD_WM8400=m
-CONFIG_MFD_WM8994=m
+# CONFIG_MFD_WM8994 is not set
# CONFIG_MFD_WM831X is not set
# CONFIG_MFD_WM8350_I2C is not set
CONFIG_MFD_PCF50633=m
--
1.7.2.3
From 52d1e96be23738ac627b3210de7cbaa11aa4a941 Mon Sep 17 00:00:00 2001
From: Sedat Dilek sedat.di...@gmail.com
Date: Mon, 13 Sep 2010 05:50:22 +0200
Subject: [PATCH] Set CONFIG_INTEL_IDLE=y
Change from 'm' to CONFIG_INTEL_IDLE=y as it is now boolean.
---
debian/config/kernelarch-x86/config |4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/debian/config/kernelarch-x86/config
b/debian/config/kernelarch-x86/config
index 009fe14..5c793c1 100644
--- a/debian/config/kernelarch-x86/config
+++ b/debian/config/kernelarch-x86/config
@@ -536,9 +536,9 @@ CONFIG_BLK_DEV_TRM290=m
CONFIG_XEN_KBDDEV_FRONTEND=y
##
-## file: drivers/indle/Kconfig
+## file: drivers/idle/Kconfig
##
-CONFIG_INTEL_IDLE=m
+CONFIG_INTEL_IDLE=y
##
## file: drivers/input/gameport/Kconfig
--
1.7.2.3
Re: Single kernel variant: Missing linux-headers-*-common package
Hi Waldi, thanks for your fast reply. Yeah, it's a bit a dirty hack, but worked-for-me. What about this solution: Putting the creation of l-h-*-common into binary-arch-flavour? OK with that one? (Untested, doing a next build with stable-queue-2.6.35 patches). - Sedat - On Wed, Aug 11, 2010 at 12:44 PM, Bastian Blank wa...@debian.org wrote: On Wed, Aug 11, 2010 at 04:16:20AM +0200, Sedat Dilek wrote: +# Single kernel variant: Create missing linux-headers-*-common package. +install-headers_$(ARCH)_$(FEATURESET)_$(FLAVOUR): install-headers_$(ARCH)_$(FEATURESET) No. This is not okay. Both targets are called independent from rules.gen. Bastian -- Another dream that failed. There's nothing sadder. -- Kirk, This side of Paradise, stardate 3417.3 diff -rdup debian/rules.real.orig debian/rules.real --- debian/rules.real.orig 2010-08-11 13:12:25.981224219 +0200 +++ debian/rules.real 2010-08-11 13:11:25.513224219 +0200 @@ -49,6 +49,8 @@ ifeq ($(DEBUG),True) endif ifeq ($(MODULES),True) binary-arch-flavour: install-headers_$(ARCH)_$(FEATURESET)_$(FLAVOUR) + # Single kernel variant: Create missing linux-headers-*-common package. + binary-arch-flavour: install-headers_$(ARCH)_$(FEATURESET) endif binary-indep: install-doc @@ -283,8 +285,9 @@ endif dh_link /usr/lib/$(PACKAGE_NAME_KBUILD)/scripts $(BASE_DIR)/scripts mkdir -p $(PACKAGE_DIR)/lib/modules/$(REAL_VERSION) - ln -s /usr/src/$(PACKAGE_NAME) $(PACKAGE_DIR)/lib/modules/$(REAL_VERSION)/build - ln -s /usr/src/$(PACKAGE_NAME_COMMON) $(PACKAGE_DIR)/lib/modules/$(REAL_VERSION)/source + # Force creation of symlinks (even if they exist). + ln -s -f /usr/src/$(PACKAGE_NAME) $(PACKAGE_DIR)/lib/modules/$(REAL_VERSION)/build + ln -s -f /usr/src/$(PACKAGE_NAME_COMMON) $(PACKAGE_DIR)/lib/modules/$(REAL_VERSION)/source install -d $(PACKAGE_DIR)/DEBIAN sed -e 's/=V/$(REAL_VERSION)/g' \
Re: Single kernel variant: Missing linux-headers-*-common package
Hi, I have uploaded my linux-2.6 and linux-kbuild-2.6 Debian packages for i386 to [1]. As single-kernel-variant I used i386_none_686. My linux-image contains 2.6.35.1 patch from upstream and stable-queue-2.6.35 patch-series from [2],[3]. I have tested with compat-wireless [4] against linux-headers packages to check if the build is OK. For easier download please use the XZ-tarball in [4] (it contains all files). Kind Regards, - Sedat - [1] http://files.iniza.org/single-kernel-variant/ [2] http://files.iniza.org/single-kernel-variant/linux-2.6/patches/ [3] http://git.kernel.org/?p=linux/kernel/git/stable/stable-queue.git;a=commit;h=fbb5de7b34810c68d1531f377458b1b7d48950ca [4] http://wireless.kernel.org/download/compat-wireless-2.6/ [5] http://files.iniza.org/single-kernel-variant/archive/ -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/aanlktimvqb=4vnmq=5ggsbypmpxn0xhjwpg6vj4nw...@mail.gmail.com
Single kernel variant: Missing linux-headers-*-common package
Hi,
while studying the kernel-buildsystem from the Debian Kernel Team and
their Wiki pages, I came accross of a known issue.
[0] says:
Problem: In this case, linux-headers-2.6.24-1+foo.1-common will be
missing. One needs to invoke the binary-arch_i386 target, which will
yield all feature sets (Xen, VServer) and flavours to be generated,
and obviously takes a lot longer.
Unfortunately, my main faster notebook has some hardware problems,
thus I am currently doing my daily job on an Intel Pentium-M Banias
1.7GHz with 1GByte RAM notebook.
This is a step back for me (Core2Duo T7200 with 2GByte RAM).
Personally, I am not interested in a full build of all i386
featuresets/flavours combinations.
Furthermore, the build of the docs deb-files took 1.5hrs, totally the
build-time was approx. 3.5hrs (following the Wiki).
EXAMPLE:
$ fakeroot make -f debian/rules.gen binary-arch_i386_none_686 binary-indep \
DEBIAN_KERNEL_JOBS=${NR_CPUS}
Not using binary-indep supresses the build of docs and source
deb-files, but also linux-base which is required by linux-image.
Building a single kernel variant creates
linux-{image,headers}-$(REAL_VERSION) packages.
(See definition in debian/rules.real 'REAL_VERSION =
$(UPSTREAMVERSION)$(ABINAME)$(LOCALVERSION)'.)
With UPSTREAMVERSION=2.6.35, ABINAME=-trunk and LOCALVERSION=686 you will get:
linux-image-2.6.35-trunk-686 and linux-headers-2.6.35-trunk-686
But linux-headers-2.6.35-trunk-686 package requires
linux-headers-2.6.35-trunk-common for installation.
Invoking install-headers_$(ARCH)_$(FEATURESET) creates the missing
binary file (see my patch).
For more Details see debian/README.Debian and [1].
Last step for a complete linux-headers installation is to apt-get
(or even to create) a linux-kbuild-2.6 package fitting to your
Linux-kernel version.
[2] has a step-by-step installation description and an official
package is being prepared.
Build instructions Single Kernel Variant:
ARCH=i386
FEATURESET=none
FLAVOUR=686
SINGLE_KERNEL_VARIANT=${ARCH}_${FEATURESET}_${FLAVOUR}
1. fakeroot make -f debian/rules debian/control-real
2. fakeroot make -f debian/rules.gen setup_${SINGLE_KERNEL_VARIANT}
3. fakeroot make -f debian/rules.gen binary-arch_${SINGLE_KERNEL_VARIANT}
4. fakeroot make -f debian/rules.real install-linux-base
5. fakeroot make -f debian/rules.real install-firmware
install-linux-base from Step #4 is required by linux-image package.
For more Details see build_linux26.sh script (it has lots of comments).
I did *not* check a full i386-arch build, so my patch might cause
unknown side-effects there.
Kind Regards,
- Sedat - (aka dileks/dileX on IRC)
[ REFERENCES ]
[0] http://lists.debian.org/debian-kernel/2008/04/msg00190.html:
linux-headers...common in single variant build
[1] http://wiki.debian.org/HowToRebuildAnOfficialDebianKernelPackage
- Building only a single kernel variant
[2] http://wiki.debian.org/HowToRebuildAnOfficialDebianKernelPackage
- The story of linux-kbuild-2.6
diff -rdup debian/rules.real.orig debian/rules.real
--- debian/rules.real.orig 2010-08-10 21:52:30.844404812 +0200
+++ debian/rules.real 2010-08-11 00:39:50.628923634 +0200
@@ -258,6 +258,15 @@ install-headers_$(ARCH)_$(FEATURESET)_$(
mkdir -p $(DIR)/arch/$(KERNEL_ARCH)/kernel
cp -a $(SOURCE_DIR)/{.config,.kernel*,Module.symvers,include} $(DIR)
cp -a $(SOURCE_DIR)/arch/$(KERNEL_ARCH)/kernel/asm-offsets.s $(DIR)/arch/$(KERNEL_ARCH)/kernel
+# Building a single kernel variant creates linux-{image,headers}-$(REAL_VERSION) packages.
+# See definition 'REAL_VERSION = $(UPSTREAMVERSION)$(ABINAME)$(LOCALVERSION)'.
+# Example: fakeroot make -f debian/rules.gen binary-arch_i386_none_686
+# With UPSTREAMVERSION=2.6.35, ABINAME=-trunk and LOCALVERSION=686 you will get:
+# linux-image-2.6.35-trunk-686 and linux-headers-2.6.35-trunk-686
+# But linux-headers-2.6.35-trunk-686 package requires linux-headers-2.6.35-trunk-common for installation.
+# Invoking install-headers_$(ARCH)_$(FEATURESET) creates the missing binary file.
+# For more Details see debian/README.Debian.
+install-headers_$(ARCH)_$(FEATURESET)_$(FLAVOUR): install-headers_$(ARCH)_$(FEATURESET)
ifeq ($(ARCH),powerpc)
if [ -f $(SOURCE_DIR)/arch/$(KERNEL_ARCH)/lib/crtsavres.o ]; then \
@@ -283,6 +292,7 @@ endif
dh_link /usr/lib/$(PACKAGE_NAME_KBUILD)/scripts $(BASE_DIR)/scripts
mkdir -p $(PACKAGE_DIR)/lib/modules/$(REAL_VERSION)
+ rm -f $(PACKAGE_DIR)/lib/modules/$(REAL_VERSION)/{build,source}
ln -s /usr/src/$(PACKAGE_NAME) $(PACKAGE_DIR)/lib/modules/$(REAL_VERSION)/build
ln -s /usr/src/$(PACKAGE_NAME_COMMON) $(PACKAGE_DIR)/lib/modules/$(REAL_VERSION)/source
build_linux26.sh
Description: Bourne shell script
Re: Single kernel variant: Missing linux-headers-*-common package
v2:
- Set install-headers_$(ARCH)_$(FEATURESET)_$(FLAVOUR):
install-headers_$(ARCH)_$(FEATURESET) *before* not after the
install-headers_$(ARCH)_$(FEATURESET)_$(FLAVOUR) block
- Replace removes via forced-symlinks
- Sedat -
On Wed, Aug 11, 2010 at 2:14 AM, Sedat Dilek sedat.di...@googlemail.com wrote:
Hi,
while studying the kernel-buildsystem from the Debian Kernel Team and
their Wiki pages, I came accross of a known issue.
[0] says:
Problem: In this case, linux-headers-2.6.24-1+foo.1-common will be
missing. One needs to invoke the binary-arch_i386 target, which will
yield all feature sets (Xen, VServer) and flavours to be generated,
and obviously takes a lot longer.
Unfortunately, my main faster notebook has some hardware problems,
thus I am currently doing my daily job on an Intel Pentium-M Banias
1.7GHz with 1GByte RAM notebook.
This is a step back for me (Core2Duo T7200 with 2GByte RAM).
Personally, I am not interested in a full build of all i386
featuresets/flavours combinations.
Furthermore, the build of the docs deb-files took 1.5hrs, totally the
build-time was approx. 3.5hrs (following the Wiki).
EXAMPLE:
$ fakeroot make -f debian/rules.gen binary-arch_i386_none_686 binary-indep \
DEBIAN_KERNEL_JOBS=${NR_CPUS}
Not using binary-indep supresses the build of docs and source
deb-files, but also linux-base which is required by linux-image.
Building a single kernel variant creates
linux-{image,headers}-$(REAL_VERSION) packages.
(See definition in debian/rules.real 'REAL_VERSION =
$(UPSTREAMVERSION)$(ABINAME)$(LOCALVERSION)'.)
With UPSTREAMVERSION=2.6.35, ABINAME=-trunk and LOCALVERSION=686 you will get:
linux-image-2.6.35-trunk-686 and linux-headers-2.6.35-trunk-686
But linux-headers-2.6.35-trunk-686 package requires
linux-headers-2.6.35-trunk-common for installation.
Invoking install-headers_$(ARCH)_$(FEATURESET) creates the missing
binary file (see my patch).
For more Details see debian/README.Debian and [1].
Last step for a complete linux-headers installation is to apt-get
(or even to create) a linux-kbuild-2.6 package fitting to your
Linux-kernel version.
[2] has a step-by-step installation description and an official
package is being prepared.
Build instructions Single Kernel Variant:
ARCH=i386
FEATURESET=none
FLAVOUR=686
SINGLE_KERNEL_VARIANT=${ARCH}_${FEATURESET}_${FLAVOUR}
1. fakeroot make -f debian/rules debian/control-real
2. fakeroot make -f debian/rules.gen setup_${SINGLE_KERNEL_VARIANT}
3. fakeroot make -f debian/rules.gen binary-arch_${SINGLE_KERNEL_VARIANT}
4. fakeroot make -f debian/rules.real install-linux-base
5. fakeroot make -f debian/rules.real install-firmware
install-linux-base from Step #4 is required by linux-image package.
For more Details see build_linux26.sh script (it has lots of comments).
I did *not* check a full i386-arch build, so my patch might cause
unknown side-effects there.
Kind Regards,
- Sedat - (aka dileks/dileX on IRC)
[ REFERENCES ]
[0] http://lists.debian.org/debian-kernel/2008/04/msg00190.html:
linux-headers...common in single variant build
[1] http://wiki.debian.org/HowToRebuildAnOfficialDebianKernelPackage
- Building only a single kernel variant
[2] http://wiki.debian.org/HowToRebuildAnOfficialDebianKernelPackage
- The story of linux-kbuild-2.6
--- debian/rules.real.orig 2010-08-11 03:47:05.876923886 +0200
+++ debian/rules.real 2010-08-11 04:01:01.444923962 +0200
@@ -239,6 +239,9 @@ install-headers_$(ARCH)_$(FEATURESET): $
+$(MAKE_SELF) install-base
+# Single kernel variant: Create missing linux-headers-*-common package.
+install-headers_$(ARCH)_$(FEATURESET)_$(FLAVOUR): install-headers_$(ARCH)_$(FEATURESET)
+
install-headers_$(ARCH)_$(FEATURESET)_$(FLAVOUR): REAL_VERSION = $(UPSTREAMVERSION)$(ABINAME)$(LOCALVERSION)
install-headers_$(ARCH)_$(FEATURESET)_$(FLAVOUR): PACKAGE_NAME = linux-headers-$(REAL_VERSION)
install-headers_$(ARCH)_$(FEATURESET)_$(FLAVOUR): PACKAGE_NAME_COMMON = linux-headers-$(UPSTREAMVERSION)$(ABINAME)-common$(LOCALVERSION_HEADERS)
@@ -283,8 +286,9 @@ endif
dh_link /usr/lib/$(PACKAGE_NAME_KBUILD)/scripts $(BASE_DIR)/scripts
mkdir -p $(PACKAGE_DIR)/lib/modules/$(REAL_VERSION)
- ln -s /usr/src/$(PACKAGE_NAME) $(PACKAGE_DIR)/lib/modules/$(REAL_VERSION)/build
- ln -s /usr/src/$(PACKAGE_NAME_COMMON) $(PACKAGE_DIR)/lib/modules/$(REAL_VERSION)/source
+ # Force creation of symlinks (even if they exist).
+ ln -s -f /usr/src/$(PACKAGE_NAME) $(PACKAGE_DIR)/lib/modules/$(REAL_VERSION)/build
+ ln -s -f /usr/src/$(PACKAGE_NAME_COMMON) $(PACKAGE_DIR)/lib/modules/$(REAL_VERSION)/source
install -d $(PACKAGE_DIR)/DEBIAN
sed -e 's/=V/$(REAL_VERSION)/g' \
Bug#574526: linux-image-2.6.32-4-amd64: iwlagn flood in dmesg: free more than tfds_in_queue
Hi Ben, there is patch pending in wireless-2.6 master GIT branch [1]: commit c8406ea8fa1adde8dc5400127281d497bbcdb84a iwlwifi: Silence tfds_in_queue message Also, it would be fine to see [2.6.32] iwlwifi: fix nfreed-- in debian's 2.6.32 kernel. Kind Regards, - Sedat - [1] http://git.kernel.org/?p=linux/kernel/git/linville/wireless-2.6.git [2] https://patchwork.kernel.org/patch/86722/ On Fri, Mar 19, 2010 at 2:15 AM, Ben Hutchings b...@decadent.org.uk wrote: On Thu, 2010-03-18 at 21:05 +0100, Cyril Brulebois wrote: Package: linux-2.6 Version: 2.6.32-10 Severity: normal Hi, it looks like the following upstream commits lead to those repeated messages, as soon as there's some traffic on the wireless network: | [ 3538.284491] iwlagn :02:00.0: free more than tfds_in_queue (1:2) | [ 3538.285838] iwlagn :02:00.0: free more than tfds_in_queue (1:2) | [ 3538.28] iwlagn :02:00.0: free more than tfds_in_queue (1:2) | [ 3538.298152] iwlagn :02:00.0: free more than tfds_in_queue (0:2) | [ 3538.299241] iwlagn :02:00.0: free more than tfds_in_queue (1:2) master : a120e912eb51e347f36c71b60a1d13af74d30e83 2.6.32.y : 44af042e42f2231579ea8ef7586d3789d198f609 | $ git show v2.6.32.9-12-g44af042 | commit 44af042e42f2231579ea8ef7586d3789d198f609 | Author: Stanislaw Gruszka sgrus...@redhat.com | Date: Fri Feb 19 15:47:33 2010 -0800 | | iwlwifi: sanity check before counting number of tfds can be free [...] Stanislaw, do you understand how this is going wrong? Do you have a fix? Ben. -- Ben Hutchings If at first you don't succeed, you're doing about average. -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/2d0a357f1003190131s447476c2o250b7ad26ee21...@mail.gmail.com
Bug#574526: [PATCH] iwlwifi: Silence tfds_in_queue message not needed
Just FYI. - Sedat - -- Forwarded message -- From: Stanislaw Gruszka sgrus...@redhat.com Date: Fri, Mar 19, 2010 at 10:37 AM Subject: Re: [PATCH] iwlwifi: Silence tfds_in_queue message To: sedat.di...@gmail.com Cc: Adel Gadllah adel.gadl...@gmail.com, reinette chatre reinette.cha...@intel.com, Greg KH gre...@suse.de, linux-wirel...@vger.kernel.org linux-wirel...@vger.kernel.org, John W. Linville linvi...@tuxdriver.com, sta...@kernel.org sta...@kernel.org, Wey-Yi Guy wey-yi.w@intel.com On Fri, Mar 19, 2010 at 10:22:14AM +0100, Sedat Dilek wrote: only to clarify: There is no need for iwlwifi: Silence tfds_in_queue message patch from wireless-2.6 GIT? I mean [2.6.32] iwlwifi: fix nfreed-- is enough for solving the mentionned problem? Yes. That was confirmed by Adel. Cheers Stanislaw -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/2d0a357f1003190255t12e70bfeub72c3887434f4...@mail.gmail.com
