Bug#845422: [PATCH] Please add BCM2835 MMC driver for Raspberry Pi 3
Source: linux Severity: wishlist Tags: patch Thank you for your work on maintaining the Linux kernel in Debian, I really appreciate it! I am interested in running Debian on the Raspberry Pi 3. As per https://github.com/anholt/linux/wiki/Raspberry-Pi-3, the mainline Linux kernel in version 4.8 includes support for the Raspberry Pi 3, so we are in pretty good shape already. However, when trying to boot linux-image-4.8.0-1-arm64 version 4.8.5-1, I noticed that the kernel can’t find any block devices (and hence no root file system). This is because the bcm2835-sdhost MMC driver has not actually made it into Linux 4.8; it is still under review: https://www.spinics.net/lists/arm-kernel/msg513433.html The driver is based on the work by Phil Elwell (from Raspberry Pi Ltd.), shipped in https://github.com/raspberrypi/linux, i.e. used by Raspbian and other distributions. Eric Anholt and Gerd Hoffmann are working upstreaming the driver. In the meantime, Gerd Hoffmann has included the current patches in his Raspberry Pi Fedora images; the source of the kernel RPMs can be found at https://www.kraxel.org/repos/rpi2/kernel-main/kernel-main-4.9-0.4.src.rpm I have applied these patches as-is to the Debian linux 4.8.5-1 package and can confirm that the resulting kernel boots on the Raspberry Pi 3 and seems to work perfectly fine. Therefore, could you please: 1. Add the attached patch series to the linux-image-4.8.0-1-arm64 build. 2. Enable the CONFIG_MMC_BCM2835=y option. Let me know whether you’d prefer me to squash these patches into a single patch file. I chose to attach the patch series as-is because that should make it easier to compare it with upstream and update it (should that become necessary). It would be brilliant to get this in before the freeze, so that people can soon run Debian stable on their Raspberry Pi 3s and get security updates :). Thank you so much! P.S.: In case you are interested, I am using the following parameters in config.txt on the FAT32 boot partition to make the Raspberry Pi bootloader load my kernel: # Switch the CPU from ARMv7 into ARMv8 (aarch64) mode arm_control=0x200 # From /usr/lib/linux-image-4.8.0-1-arm64/broadcom/bcm2837-rpi-3-b.dtb device_tree=bcm2837-rpi-3-b.dtb kernel=vmlinuz-4.8.0-1-arm64 # For details on the initramfs directive, see # https://www.raspberrypi.org/forums/viewtopic.php?f=63&t=10532 initramfs initrd.img-4.8.0-1-arm64 -- System Information: Debian Release: stretch/sid APT prefers testing APT policy: (990, 'testing'), (500, 'unstable') Architecture: amd64 (x86_64) Foreign Architectures: i386, armel, mipsel Kernel: Linux 4.6.0-1-amd64 (SMP w/8 CPU cores) Locale: LANG=de_DE.UTF-8, LC_CTYPE=de_DE.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) >From 48ec9c44ea4d650d0d673c3a9df918152007e8e7 Mon Sep 17 00:00:00 2001 From: Eric Anholt Date: Mon, 22 Feb 2016 12:44:55 -0800 Subject: [PATCH 23/43] mmc: bcm2835-sdhost: Add new driver for the internal SD controller. The 2835 has two SD controllers: The Arasan SDHCI controller that we currently use, and a custom SD controller. The custom one runs faster The code was originally written by Phil Elwell in the downstream Rasbperry Pi tree, and I did a major cleanup on it (+319, -707 lines out of the original 2055) for inclusion. Signed-off-by: Eric Anholt --- drivers/mmc/host/Kconfig |9 + drivers/mmc/host/Makefile |1 + drivers/mmc/host/bcm2835-sdhost.c | 1666 + 3 files changed, 1676 insertions(+) create mode 100644 drivers/mmc/host/bcm2835-sdhost.c diff --git a/drivers/mmc/host/Kconfig b/drivers/mmc/host/Kconfig index 5274f50..e928160 100644 --- a/drivers/mmc/host/Kconfig +++ b/drivers/mmc/host/Kconfig @@ -769,6 +769,15 @@ config MMC_TOSHIBA_PCI depends on PCI help +config MMC_BCM2835_SDHOST + tristate "platform support for the BCM2835 SDHOST MMC Controller" + depends on ARCH_BCM2835 || COMPILE_TEST + help + This selects the BCM2835 SDHOST MMC controller. If you have a BCM2835 + platform with SD or MMC devices, say Y or M here. + + If unsure, say N. + config MMC_MTK tristate "MediaTek SD/MMC Card Interface support" depends on HAS_DMA diff --git a/drivers/mmc/host/Makefile b/drivers/mmc/host/Makefile index e2bdaaf4..2fb3f36 100644 --- a/drivers/mmc/host/Makefile +++ b/drivers/mmc/host/Makefile @@ -57,6 +57,7 @@ obj-$(CONFIG_MMC_MOXART) += moxart-mmc.o obj-$(CONFIG_MMC_SUNXI) += sunxi-mmc.o obj-$(CONFIG_MMC_USDHI6ROL0) += usdhi6rol0.o obj-$(CONFIG_MMC_TOSHIBA_PCI) += toshsd.o +obj-$(CONFIG_MMC_BCM2835_SDHOST) += bcm2835-sdhost.o obj-$(CONFIG_MMC_REALTEK_PCI) += rtsx_pci_sdmmc.o obj-$(CONFIG_MMC_REALTEK_USB) += rtsx_usb_sdmmc.o diff --git a/drivers/mmc/host/bcm2835-sdhost.c b/drivers/mmc/host/bcm2835-sdhost.c new file mode 100644 index 000..cafe6ad --- /dev/null +++ b/drivers/mmc/host/bcm2835-sdhost.c @@ -0,0 +1,1666 @@ +/* + * BCM2835 SD host driver. + * + * Aut
Processed: your mail
Processing commands for cont...@bugs.debian.org: > reassign 839381 linux Bug #839381 [xserver-xorg-video-amdgpu] xserver-xorg-video-amdgpu: very long Xorg initilization time in recent kernels Bug reassigned from package 'xserver-xorg-video-amdgpu' to 'linux'. No longer marked as found in versions xserver-xorg-video-amdgpu/1.1.2-1. Ignoring request to alter fixed versions of bug #839381 to the same values previously set > End of message, stopping processing here. Please contact me if you need assistance. -- 839381: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=839381 Debian Bug Tracking System Contact ow...@bugs.debian.org with problems
Processed: your mail
Processing commands for cont...@bugs.debian.org: > tags 839381 +patch Bug #839381 [linux] xserver-xorg-video-amdgpu: very long Xorg initilization time in recent kernels Added tag(s) patch. > End of message, stopping processing here. Please contact me if you need assistance. -- 839381: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=839381 Debian Bug Tracking System Contact ow...@bugs.debian.org with problems
Re: cross building linux-image packages
> From: Ben Hutchings > > > > In particular trying to do > > > > fakeroot make -f debian/rules.gen binary-arch_armel_none > > > > but that doesn't seem work (gcc-5 command not found?) under jessie - > > now looking at under stretch... > > > > But I think I am straying too far from the beaten track and hitting > > lots of problems - hence the question about where information on what > > people are normally doing to cross build the kernel using debian > > packages. > > Using the stretch branch you should use something like: > > dpkg-buildpackage -Pcross,nopython -aarmel -B -uc -us > > I got that working (except with armhf as the target) a few months back. > > You can also add 'notools' to the -P option if you don't need the > userland packages; this will reduce the build-dependencies. > > Ben. Thank-you very much Ben - got that approach to work (eventually :-) I tried: dpkg-buildpackage -Pcross,nopython -aarmel -B -uc -us First I got several dependancy complaints that were not valid. The components that dpkg-buildpackage complained about were apparently already installed according to apt-get install libssl-dev libglib2.0-dev libudev-dev libwrap0-dev libpci-dev gcc-5-arm-linux-gnueabi:native So I tried it with -d and got *lot* of compiling ending with the failures and requiring some missing parts to be installed... * missing arm-linux-gnueabi-gcc - installed the package gcc-arm-linux-gnueabi * missing openssl/opensslconf.h: - installed libssl-dev:armel * missing libpci - installed libpci-dev:armel * missing libwrap - installed libwrap0-dev:armel Glad I found out about the -nc option to avoid the lengthy rebuilding from scratch... * Hmm apparently the complaints were valid - but for the armel packages! Thanks again. Andrew
Bug#845382: more boot BUG logs
I added netconsole into /boot/grub/grub.cfg and did more attempts to boot. Logs are attached in the archive. zito@bobek:~$ fgrep -m 1 BUG radeon-kms-messages* radeon-kms-messages.txt:[ 64.148004] NMI watchdog: BUG: soft lockup - CPU#1 stuck for 22s! [kworker/1:2:51] radeon-kms-messages1.txt:[ 14.147686] BUG: unable to handle kernel NULL pointer dereference at 006c radeon-kms-messages2.txt:[ 15.379376] BUG: unable to handle kernel NULL pointer dereference at 006c radeon-kms-messages3.txt:[ 13.890714] BUG: unable to handle kernel NULL pointer dereference at 006c radeon-kms-messages4.txt:[ 14.579563] BUG: unable to handle kernel NULL pointer dereference at 006c radeon-kms-messages5.txt:[ 15.651245] BUG: unable to handle kernel NULL pointer dereference at 006c radeon-kms-messages6.txt:[ 48.148004] NMI watchdog: BUG: soft lockup - CPU#1 stuck for 22s! [smartd:457] radeon-kms-messages7.txt:[ 15.472800] BUG: unable to handle kernel NULL pointer dereference at 006c radeon-kms-messages8.txt:[ 14.282438] BUG: unable to handle kernel NULL pointer dereference at 006c radeon-kms-messages9.txt:[ 14.860173] BUG: unable to handle kernel NULL pointer dereference at 006c There is NULL pointer dereference in most cases. -- Zito radeon-kms-messages.tar.xz Description: application/xz
Processed: retitle 845302 to used mount implementations are incomplete, fails to mount /usr
Processing commands for cont...@bugs.debian.org: > retitle 845302 used mount implementations are incomplete, fails to mount /usr Bug #845302 [initramfs-tools] systemd: 232-6:Failed to boot, makes kernel panic when starting /sbin/init. Changed Bug title to 'used mount implementations are incomplete, fails to mount /usr' from 'systemd: 232-6:Failed to boot, makes kernel panic when starting /sbin/init.'. > thanks Stopping processing here. Please contact me if you need assistance. -- 845302: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=845302 Debian Bug Tracking System Contact ow...@bugs.debian.org with problems
Re: Bug#845302: systemd: 232-6:Failed to boot, makes kernel panic when starting /sbin/init.
Control: reassign -1 initramfs-tools 0.125 Am 23.11.2016 um 11:25 schrieb K.Ohta: > Not mounted /usr and made kernel panic still longer sleep. > > Then booting with "init=/bin/bash" , check dmesg (attached). > Below message was wrote when trying to mount /usr : >> [ 15.958645] EXT4-fs (sde3): Unrecognized mount option "auto" or missing >> value > > If this message is correct, mount program included by initramfs > has not recognized "auto" (or some another) option(s), regardless > /bin/mount (or /sbin/mount/ext4) recognize this option(s) :-( I'm going to reassign this to initramfs-tools as it uses mount implementations from either klibc-utils or busybox which are both incomplete and as a result, fail to mount /usr. I decided against reassigning this to busybox and klibc-utils. I think it's preferable, if initramfs-tools simply uses the mount implementation from the real systemd, i.e. util-linux. This guarantees that we won't run into such problems again in the future. The mount utility is tiny, all its library dependencies are already in the initramfs. Rebuilding the initramfs with mount from u-l increases the size by 17K. We would also have to make sure to use mount --move in /usr/share/initramfs-tools/init for /run, /sys and /proc Ben, what's your take on this? Regards, Michael -- Why is it that all of the instruments seeking intelligent life in the universe are pointed away from Earth? signature.asc Description: OpenPGP digital signature
Processed: Re: Bug#845302: systemd: 232-6:Failed to boot, makes kernel panic when starting /sbin/init.
Processing control commands: > reassign -1 initramfs-tools 0.125 Bug #845302 [systemd] systemd: 232-6:Failed to boot, makes kernel panic when starting /sbin/init. Bug reassigned from package 'systemd' to 'initramfs-tools'. No longer marked as found in versions systemd/232-6. Ignoring request to alter fixed versions of bug #845302 to the same values previously set Bug #845302 [initramfs-tools] systemd: 232-6:Failed to boot, makes kernel panic when starting /sbin/init. Marked as found in versions initramfs-tools/0.125. -- 845302: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=845302 Debian Bug Tracking System Contact ow...@bugs.debian.org with problems
Processed: tagging 845302
Processing commands for cont...@bugs.debian.org: > tags 845302 - moreinfo unreproducible confirmed Bug #845302 [initramfs-tools] used mount implementations are incomplete, fails to mount /usr Removed tag(s) moreinfo and unreproducible. > thanks Stopping processing here. Please contact me if you need assistance. -- 845302: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=845302 Debian Bug Tracking System Contact ow...@bugs.debian.org with problems
Bug#845302: systemd: 232-6:Failed to boot, makes kernel panic when starting /sbin/init.
Am 23.11.2016 um 14:30 schrieb Michael Biebl:
> I'm going to reassign this to initramfs-tools as it uses mount
> implementations from either klibc-utils or busybox which are both
> incomplete and as a result, fail to mount /usr.
>
> I decided against reassigning this to busybox and klibc-utils.
> I think it's preferable, if initramfs-tools simply uses the mount
> implementation from the real systemd, i.e. util-linux. This guarantees
> that we won't run into such problems again in the future.
>
> The mount utility is tiny, all its library dependencies are already in
> the initramfs. Rebuilding the initramfs with mount from u-l increases
> the size by 17K.
>
> We would also have to make sure to use mount --move in
> /usr/share/initramfs-tools/init for /run, /sys and /proc
>
> Ben, what's your take on this?
Something like the attached patch should do the trick.
I wonder though, if we should use the other tools from mount as well,
like umount, swapon/off
Feedback welcome.
--
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?
From dd370d9b7242624f67e6263177997446e01705ba Mon Sep 17 00:00:00 2001
From: Michael Biebl
Date: Wed, 23 Nov 2016 15:54:38 +0100
Subject: [PATCH] Use mount implementation from util-linux
It's also used on the real system and seems to be the only complete
implementation.
Closes: #845302
---
debian/control | 2 +-
hooks/mount| 23 +++
init | 6 +++---
3 files changed, 27 insertions(+), 4 deletions(-)
create mode 100755 hooks/mount
diff --git a/debian/control b/debian/control
index fc9cb7b..813fde5 100644
--- a/debian/control
+++ b/debian/control
@@ -25,7 +25,7 @@ Package: initramfs-tools-core
Architecture: all
Multi-Arch: foreign
Recommends: ${busybox:Recommends}
-Depends: klibc-utils (>= 2.0.4-8~), cpio, kmod | module-init-tools, udev, ${misc:Depends}
+Depends: klibc-utils (>= 2.0.4-8~), mount, cpio, kmod | module-init-tools, udev, ${misc:Depends}
Suggests: bash-completion
Breaks: initramfs-tools (<< 0.121~)
Replaces: initramfs-tools (<< 0.121~)
diff --git a/hooks/mount b/hooks/mount
new file mode 100755
index 000..1464533
--- /dev/null
+++ b/hooks/mount
@@ -0,0 +1,23 @@
+#!/bin/sh
+
+PREREQ="klibc-utils"
+
+prereqs()
+{
+ echo "$PREREQ"
+}
+
+case $1 in
+prereqs)
+ prereqs
+ exit 0
+ ;;
+esac
+
+. /usr/share/initramfs-tools/hook-functions
+
+# We pulled in klibc-utils, to make sure we run after it,
+# so we can replace the symlink
+rm -f ${DESTDIR}/bin/mount
+copy_exec /bin/mount
+
diff --git a/init b/init
index 9552e5c..886a55e 100755
--- a/init
+++ b/init
@@ -248,7 +248,7 @@ run_scripts /scripts/init-bottom
[ "$quiet" != "y" ] && log_end_msg
# Move /run to the root
-mount -n -o move /run ${rootmnt}/run
+mount -n --move /run ${rootmnt}/run
validate_init() {
run-init -n "${rootmnt}" "${1}"
@@ -299,8 +299,8 @@ unset forcefsck
unset fsckfix
# Move virtual filesystems over to the real filesystem
-mount -n -o move /sys ${rootmnt}/sys
-mount -n -o move /proc ${rootmnt}/proc
+mount -n --move /sys ${rootmnt}/sys
+mount -n --move /proc ${rootmnt}/proc
# Chain to real filesystem
exec run-init ${drop_caps} ${rootmnt} ${init} "$@" <${rootmnt}/dev/console >${rootmnt}/dev/console 2>&1
--
2.10.2
signature.asc
Description: OpenPGP digital signature
Processed: affects 845500
Processing commands for cont...@bugs.debian.org: > affects 845500 nftables Bug #845500 [linux] nftables: notrack target fails with No such file or directory Added indication that 845500 affects nftables > thanks Stopping processing here. Please contact me if you need assistance. -- 845500: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=845500 Debian Bug Tracking System Contact ow...@bugs.debian.org with problems
Processed: Re: nftables: notrack target fails with No such file or directory
Processing control commands: > reassign -1 linux 4.9~rc5-1~exp1 Bug #845500 [nftables] nftables: notrack target fails with No such file or directory Bug reassigned from package 'nftables' to 'linux'. No longer marked as found in versions nftables/0.6+snapshot20161117-2. Ignoring request to alter fixed versions of bug #845500 to the same values previously set Bug #845500 [linux] nftables: notrack target fails with No such file or directory There is no source info for the package 'linux' at version '4.9~rc5-1~exp1' with architecture '' Unable to make a source version for version '4.9~rc5-1~exp1' Marked as found in versions 4.9~rc5-1~exp1. -- 845500: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=845500 Debian Bug Tracking System Contact ow...@bugs.debian.org with problems
Bug#845500: nftables: notrack target fails with No such file or directory
On 24.11.2016 01:34, Peter Colberg wrote: > Assuming 4.9 becomes the stretch kernel, could you backport the patch? According to https://lists.debian.org/debian-devel-announce/2016/03/msg0.html it will be 4.10. Greets jre
Bug#845500: nftables: notrack target fails with No such file or directory
On Wed, Nov 23, 2016 at 07:34:42PM -0500, Peter Colberg wrote: > Assuming 4.9 becomes the stretch kernel, could you backport the patch? The same applies to kernel support for the "fib" expression that may be used for reverse path filtering (analogous to iptables rp_filter). https://git.kernel.org/cgit/linux/kernel/git/pablo/nf-next.git/commit?id=f6d0cbcf09c506b9b022df8f9d7693a7cec3c732 That patch is more extensive and there are many more commits needed to sync nftables kernel support with userspace. Backporting does not make much sense. I am crossing fingers for 4.10 making it into stretch. Peter
Bug#845500: nftables: notrack target fails with No such file or directory
On Thu, Nov 24, 2016 at 01:55:01AM +0100, Jens Reyer wrote: > According to > https://lists.debian.org/debian-devel-announce/2016/03/msg0.html it > will be 4.10. That would be great. After the recent announcement that 4.9 will probably be the next LTS kernel I assumed that the same version would also be shipped with stretch. http://kroah.com/log/blog/2016/09/06/4-dot-9-equals-equals-next-lts-kernel/ Peter
Processed: retitle 845302 to mount implementations from busybox and klibc-utils incomplete, fail to mount /usr when using auto option
Processing commands for cont...@bugs.debian.org: > retitle 845302 mount implementations from busybox and klibc-utils incomplete, > fail to mount /usr when using auto option Bug #845302 [initramfs-tools] used mount implementations are incomplete, fails to mount /usr Changed Bug title to 'mount implementations from busybox and klibc-utils incomplete, fail to mount /usr when using auto option' from 'used mount implementations are incomplete, fails to mount /usr'. > thanks Stopping processing here. Please contact me if you need assistance. -- 845302: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=845302 Debian Bug Tracking System Contact ow...@bugs.debian.org with problems
Bug#838887: please dont bomb out on nofail option
On Mon, 26 Sep 2016 07:02:01 +0200 Marc Haber wrote: > Package: initramfs-tools > Severity: wishlist > > Hi, > > when I looked for the last time (a few months ago), fstab entries > parsed by initramfs-tools or in initramfs MUST NOT contain the > "nofail" option. This is kind of surprising since some other fstab > entries MUST contain the "nofail" option to avoid systemd from bombing > out on boot. > > This inconsistence makey it necessary to think in which lines to put > the nofail option and in which lines not. > > Please consider tweaking initramfs-tools so that "nofail" can be in > all lines, with it being ignored in initramfs. This looks like a duplicate of #845302. In this bug report a user had the "auto" option for /usr in /etc/fstab. The busybox and klibc-utils mount implementations did not cope with this and failed to mount /usr as a result In https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=845302#161 I have a proposed patch which uses the mount implmentation from util-linux, which is also used on the real system. This should avoid issues like Mark or the bug reporter from #845302 are encountering. -- Why is it that all of the instruments seeking intelligent life in the universe are pointed away from Earth? signature.asc Description: OpenPGP digital signature
Bug#845500: nftables: notrack target fails with No such file or directory
On 24 November 2016 at 01:34, Peter Colberg wrote: > > While the nftables package in Debian stretch will support notrack, the > corresponding kernel support was committed after the 4.9 merge window: > > https://git.kernel.org/cgit/linux/kernel/git/pablo/nf-next.git/commit/net/netfilter/nft_ct.c?id=254432613c588640f8b8b5c3641a3c27bbe14688 > > Assuming 4.9 becomes the stretch kernel, could you backport the patch? Debian stretch will include linux 4.10 [0], so no problem. [0] https://lists.debian.org/debian-devel-announce/2016/03/msg0.html
