raspi-nolpae flavor is pointless nowadays

Bug #2023359 reported by Juerg Haefliger
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
linux-raspi (Ubuntu)
Fix Released
Undecided
Unassigned
Focal
Invalid
Undecided
Unassigned
Jammy
Fix Released
Undecided
Unassigned
Kinetic
Won't Fix
Undecided
Unassigned
Lunar
Fix Released
Undecided
Unassigned
Mantic
Fix Released
Undecided
Unassigned

Bug Description

[Impact]

linux-raspi 5.4 in Focal was the first kernel to support Pi 4B, which requires LPAE to be enabled for armhf. Since we currently don't support board-specific kernels but follow a one-kernel-fits-all strategy, the LPAE kernel is also used on Pi 3 and 2 boards. Back then, running an armhf LPAE kernel on a Pi 3/2 resulted in additional kernel memory consumption of ~70MB compared to arm64 and non-LPAE kernels. For that reason I decided to introduce a raspi-nolpae flavor which can be used on Pi 3 and 2 (only). That flavor was never officially announced nor installed automatically anywhere. A user has to install it manually. We never had any complaints about armhf kernel memory consumption so never directed anybody at installing this flavor. Therefore I'm assuming it's not being used...

Turns out that the early 5.4 kernels were broken and incorrectly initialized a 64MB SWIOTLB buffer which is the reason for the above additional memory consumption. This was fixed here: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=fcf044891c84e38fc90eb736b818781bccf94e38

With this commit, the memory consumption of both the armhf raspi and raspi-nolpae flavor are pretty much on par [1] and there is really no reason anymore to keep the raspi-nolpae flavor around. So let's get rid of it and introduce a transitional package that replaces installed raspi-nolpae packages with raspi packages.

[Test Case]

Install raspi-nolpae flavor and upgrade to the new kernel and verify that the raspi flavor is installed instead.

[Where Problems Could Occur]

Only users that have the raspi-nolpae flavor installed are affected. Upgrades, i.e., transitions from raspi-nolpae to raspi could go wrong. It's not expected that there is a user visible difference between running a current raspi-nolpae and raspi kernel flavor.

[1]
arm64 raspi -- Memory: 815528K/970752K available (13376K kernel code, 2472K rwdata, 4296K rodata, 5504K init, 850K bss, 89688K reserved, 65536K cma-reserved)
armhf raspi -- Memory: 835732K/970752K available (12288K kernel code, 1643K rwdata, 3692K rodata, 2048K init, 533K bss, 69484K reserved, 65536K cma-reserved, 118784K highmem)
armhf raspi-nolpae -- Memory: 836620K/970752K available (12288K kernel code, 1633K rwdata, 3676K rodata, 1024K init, 511K bss, 68596K reserved, 65536K cma-reserved)

Juerg Haefliger (juergh)
Changed in linux-raspi (Ubuntu Kinetic):
status: New → Won't Fix
status: Won't Fix → New
summary: - -nolpae flavor is pointless nowadays
+ raspi-nolpae flavor is pointless nowadays
Juerg Haefliger (juergh)
Changed in linux-raspi (Ubuntu Focal):
status: New → Invalid
Juerg Haefliger (juergh)
tags: added: kern-7041
Juerg Haefliger (juergh)
description: updated
description: updated
Juerg Haefliger (juergh)
Changed in linux-raspi (Ubuntu Lunar):
status: New → Fix Committed
Changed in linux-raspi (Ubuntu Kinetic):
status: New → Fix Committed
Changed in linux-raspi (Ubuntu Jammy):
status: New → Fix Committed
Revision history for this message
Ubuntu SRU Bot (ubuntu-sru-bot) wrote : Autopkgtest regression report (linux-meta-raspi/6.2.0.1008.11)

All autopkgtests for the newly accepted linux-meta-raspi (6.2.0.1008.11) for lunar have finished running.
The following regressions have been reported in tests triggered by the package:

bbswitch/0.8-14ubuntu1 (arm64)
dpdk-kmods/unknown (arm64)
linux-apfs-rw/unknown (arm64)
zfs-linux/2.1.9-2ubuntu1.1 (arm64)

Please visit the excuses page listed below and investigate the failures, proceeding afterwards as per the StableReleaseUpdates policy regarding autopkgtest regressions [1].

https://people.canonical.com/~ubuntu-archive/proposed-migration/lunar/update_excuses.html#linux-meta-raspi

[1] https://wiki.ubuntu.com/StableReleaseUpdates#Autopkgtest_Regressions

Thank you!

Revision history for this message
Ubuntu Kernel Bot (ubuntu-kernel-bot) wrote :

This bug is awaiting verification that the linux-raspi/6.2.0-1008.10 kernel in -proposed solves the problem. Please test the kernel and update this bug with the results. If the problem is solved, change the tag 'verification-needed-lunar' to 'verification-done-lunar'. If the problem still exists, change the tag 'verification-needed-lunar' to 'verification-failed-lunar'.

If verification is not done by 5 working days from today, this fix will be dropped from the source code, and this bug will be closed.

See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Thank you!

tags: added: kernel-spammed-lunar-linux-raspi verification-needed-lunar
Revision history for this message
Manuel Diewald (diewald) wrote :

I verified that upgrading from the latest raspi-nolpae flavor (6.2.0-1007-raspi-nolpae) that was installed on a fresh server armhf lunar image based system to the kernel in proposed does indeed install the regular raspi flavor (6.2.0-1008-raspi).

tags: added: verification-done-lunar
removed: verification-needed-lunar
Revision history for this message
Launchpad Janitor (janitor) wrote :
Download full text (10.9 KiB)

This bug was fixed in the package linux-raspi - 6.2.0-1008.10

---------------
linux-raspi (6.2.0-1008.10) lunar; urgency=medium

  * lunar/linux-raspi: 6.2.0-1008.10 -proposed tracker (LP: #2025455)

  * Packaging resync (LP: #1786013)
    - [Packaging] update Ubuntu.md
    - [Packaging] update annotations scripts

  * raspi-nolpae flavor is pointless nowadays (LP: #2023359)
    - [Packaging] raspi: Remove raspi-nolpae armhf flavor

  * Pi Camera 3 not working with Lunar (LP: #2024434)
    - [Config] raspi: Set VIDEO_DW9807_VCM=m

  * i2c devices missing in lunar (LP: #2019320)
    - [Config] raspi: Set I2C_CHARDEV=y

  [ Ubuntu: 6.2.0-25.25 ]

  * lunar/linux: 6.2.0-25.25 -proposed tracker (LP: #2024167)
  * ftrace in ubuntu_kernel_selftests failed with "check if duplicate events are
    caught" on J-5.15 P9 / J-kvm / L-kvm (LP: #1977827)
    - SAUCE: selftests/ftrace: Add test dependency
  * Add microphone support of the front headphone port on P3 Tower
    (LP: #2023650)
    - ALSA: hda/realtek: Add Lenovo P3 Tower platform
  * Add audio support for ThinkPad P1 Gen 6 and Z16 Gen 2 (LP: #2023539)
    - ALSA: hda/realtek: Add quirk for ThinkPad P1 Gen 6
  * Fix Disable thunderbolt clx make edp-monitor garbage while moving the
    touchpad (LP: #2023004)
    - drm/i915: Use 18 fast wake AUX sync len
  * Fix Monitor lost after replug WD19TBS to SUT port with VGA/DVI to type-C
    dongle (LP: #2021949)
    - thunderbolt: Increase timeout of DP OUT adapter handshake
    - thunderbolt: Do not touch CL state configuration during discovery
    - thunderbolt: Increase DisplayPort Connection Manager handshake timeout
  * Enable Tracing Configs for OSNOISE and TIMERLAT (LP: #2018591)
    - [Config] Enable OSNOISE_TRACER and TIMERLAT_TRACER configs
  * Fix only reach PC3 when ethernet is plugged r8169 (LP: #1946433)
    - r8169: use spinlock to protect mac ocp register access
    - r8169: use spinlock to protect access to registers Config2 and Config5
    - r8169: enable cfg9346 config register access in atomic context
    - r8169: prepare rtl_hw_aspm_clkreq_enable for usage in atomic context
    - r8169: disable ASPM during NAPI poll
    - r8169: remove ASPM restrictions now that ASPM is disabled during NAPI poll
  * introduce do_lib_rust=true|false to enable/disable linux-lib-rust package
    (LP: #2021605)
    - [Packaging] introduce do_lib_rust and enable it only on generic amd64
  * System either hang with black screen or rebooted on entering suspend on AMD
    Ryzen 9 PRO 7940HS w/ Radeon 780M Graphics (LP: #2020685)
    - drm/amdgpu: refine get gpu clock counter method
    - drm/amdgpu/gfx11: update gpu_clock_counter logic
  * generate linux-lib-rust only on amd64 (LP: #2020356)
    - [Packaging] generate linux-lib-rust only on amd64
  * No HDMI/DP audio output on dock(Nvidia GPU) (LP: #2020062)
    - ALSA: hda: Add NVIDIA codec IDs a3 through a7 to patch table
  * Add support for mdev_set_iommu_device() kABI in Ubuntu 22.10 kernel
    (LP: #1988806)
    - SAUCE: Add mdev_set_iommu_device() kABI.
  * Enable audio LEDs on HP laptops (LP: #2019915)
    - ALSA: hda/realtek: Fix mute and micmute LEDs for an HP laptop
    - ALSA: hda/realtek: Fix m...

Changed in linux-raspi (Ubuntu Lunar):
status: Fix Committed → Fix Released
Revision history for this message
Ubuntu Kernel Bot (ubuntu-kernel-bot) wrote :

This bug is awaiting verification that the linux-raspi/5.15.0-1035.38 kernel in -proposed solves the problem. Please test the kernel and update this bug with the results. If the problem is solved, change the tag 'verification-needed-jammy' to 'verification-done-jammy'. If the problem still exists, change the tag 'verification-needed-jammy' to 'verification-failed-jammy'.

If verification is not done by 5 working days from today, this fix will be dropped from the source code, and this bug will be closed.

See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Thank you!

tags: added: kernel-spammed-jammy-linux-raspi verification-needed-jammy
Revision history for this message
Ubuntu SRU Bot (ubuntu-sru-bot) wrote : Autopkgtest regression report (linux-meta-raspi/5.15.0.1035.33)

All autopkgtests for the newly accepted linux-meta-raspi (5.15.0.1035.33) for jammy have finished running.
The following regressions have been reported in tests triggered by the package:

adv-17v35x/unknown (arm64)
backport-iwlwifi-dkms/unknown (arm64)
bbswitch/unknown (arm64)
dahdi-linux/unknown (arm64)
digimend-dkms/unknown (arm64)
dm-writeboost/unknown (arm64)
dpdk-kmods/unknown (arm64)
fwts/unknown (arm64)
glibc/unknown (arm64)
jool/unknown (arm64)
langford/unknown (arm64)
lime-forensics/unknown (arm64)
nat-rtsp/unknown (arm64)
nvidia-graphics-drivers-390/390.157-0ubuntu0.22.04.2 (armhf)
nvidia-graphics-drivers-525/unknown (arm64)
wireguard/unknown (amd64)
xtables-addons/unknown (arm64)
zfs-linux/unknown (arm64)

Please visit the excuses page listed below and investigate the failures, proceeding afterwards as per the StableReleaseUpdates policy regarding autopkgtest regressions [1].

https://people.canonical.com/~ubuntu-archive/proposed-migration/jammy/update_excuses.html#linux-meta-raspi

[1] https://wiki.ubuntu.com/StableReleaseUpdates#Autopkgtest_Regressions

Thank you!

Revision history for this message
Juerg Haefliger (juergh) wrote :

Installed fresh jammy armhf image and replaced regular raspi kernel with with raspi-nolpae kernel. Added proposed PPA and did an upgrade. Verified that regular raspi flavor from proposed PPA is now installed.

tags: added: verification-done-jammy
removed: verification-needed-jammy
Revision history for this message
Launchpad Janitor (janitor) wrote :
Download full text (84.0 KiB)

This bug was fixed in the package linux-raspi - 5.15.0-1035.38

---------------
linux-raspi (5.15.0-1035.38) jammy; urgency=medium

  * jammy/linux-raspi: 5.15.0-1035.38 -proposed tracker (LP: #2026526)

  * Packaging resync (LP: #1786013)
    - [Packaging] resync getabis

  * raspi-nolpae flavor is pointless nowadays (LP: #2023359)
    - [Packaging] raspi: Remove raspi-nolpae armhf flavor

  [ Ubuntu: 5.15.0-79.86 ]

  * jammy/linux: 5.15.0-79.86 -proposed tracker (LP: #2026531)
  * Jammy update: v5.15.111 upstream stable release (LP: #2025095)
    - ASOC: Intel: sof_sdw: add quirk for Intel 'Rooks County' NUC M15
    - ASoC: soc-pcm: fix hw->formats cleared by soc_pcm_hw_init() for dpcm
    - x86/hyperv: Block root partition functionality in a Confidential VM
    - iio: adc: palmas_gpadc: fix NULL dereference on rmmod
    - ASoC: Intel: bytcr_rt5640: Add quirk for the Acer Iconia One 7 B1-750
    - selftests mount: Fix mount_setattr_test builds failed
    - asm-generic/io.h: suppress endianness warnings for readq() and writeq()
    - x86/cpu: Add model number for Intel Arrow Lake processor
    - wireguard: timers: cast enum limits members to int in prints
    - wifi: mt76: mt7921e: Set memory space enable in PCI_COMMAND if unset
    - arm64: Always load shadow stack pointer directly from the task struct
    - arm64: Stash shadow stack pointer in the task struct on interrupt
    - PCI: pciehp: Fix AB-BA deadlock between reset_lock and device_lock
    - PCI: qcom: Fix the incorrect register usage in v2.7.0 config
    - IMA: allow/fix UML builds
    - USB: dwc3: fix runtime pm imbalance on probe errors
    - USB: dwc3: fix runtime pm imbalance on unbind
    - hwmon: (k10temp) Check range scale when CUR_TEMP register is read-write
    - hwmon: (adt7475) Use device_property APIs when configuring polarity
    - posix-cpu-timers: Implement the missing timer_wait_running callback
    - blk-mq: release crypto keyslot before reporting I/O complete
    - blk-crypto: make blk_crypto_evict_key() return void
    - blk-crypto: make blk_crypto_evict_key() more robust
    - ext4: use ext4_journal_start/stop for fast commit transactions
    - staging: iio: resolver: ads1210: fix config mode
    - tty: Prevent writing chars during tcsetattr TCSADRAIN/FLUSH
    - xhci: fix debugfs register accesses while suspended
    - tick/nohz: Fix cpu_is_hotpluggable() by checking with nohz subsystem
    - MIPS: fw: Allow firmware to pass a empty env
    - ipmi:ssif: Add send_retries increment
    - ipmi: fix SSIF not responding under certain cond.
    - kheaders: Use array declaration instead of char
    - wifi: mt76: add missing locking to protect against concurrent rx/status
      calls
    - pwm: meson: Fix axg ao mux parents
    - pwm: meson: Fix g12a ao clk81 name
    - soundwire: qcom: correct setting ignore bit on v1.5.1
    - pinctrl: qcom: lpass-lpi: set output value before enabling output
    - ring-buffer: Sync IRQ works before buffer destruction
    - crypto: api - Demote BUG_ON() in crypto_unregister_alg() to a WARN_ON()
    - crypto: safexcel - Cleanup ring IRQ workqueues on load failure
    - rcu: Avoid stack overflow due to __rcu_irq_enter_check_tick() bei...

Changed in linux-raspi (Ubuntu Jammy):
status: Fix Committed → Fix Released
Revision history for this message
Juerg Haefliger (juergh) wrote :

Kinetic is EOL.

Changed in linux-raspi (Ubuntu Kinetic):
status: Fix Committed → Won't Fix
Revision history for this message
Launchpad Janitor (janitor) wrote :
Download full text (9.8 KiB)

This bug was fixed in the package linux-raspi - 6.5.0-1002.2

---------------
linux-raspi (6.5.0-1002.2) mantic; urgency=medium

  * mantic/linux-raspi: 6.5.0-1002.2 -proposed tracker (LP: #2033637)

  * Packaging resync (LP: #1786013)
    - [Packaging] update update.conf

  * Miscellaneous Ubuntu changes
    - [Config] raspi: Add notes to configs that differ from master
    - [Config] raspi: Set CONFIG_ARM_ERRATA_<foo>=n
    - [Config] raspi: Set PREEMPT_DYNAMIC=y
    - [Config] raspi: Set STAGING_MEDIA_DEPRECATED=n
    - [Config] raspi: Set EFI=n
    - [Config] raspi: Set POWER_RESET_GPIO_RESTART=n
    - [Config] raspi: Set HID_BPF=n
    - [Config] raspi: Set COMMON_CLK_FIXED_MMIO=n
    - [Config] raspi: Set PWM_STMPE=n
    - [Config] raspi: Set LEDS_SYSCON=n
    - [Packaging] raspi: Include bcm2835_smi module in linux-modules
    - SAUCE: (no-up) Disable FIQ split-transaction FSM in dwc_otg

  * Miscellaneous upstream changes
    - drivers: media: imx296: Updated imx296 driver for external trigger
    - char: broadcom: vc_mem: Fix preprocessor conditional
    - drivers: dwc_otg: Fix fallthrough warnings
    - vc04_services/vc-sm-cma: Switch one-bit bitfields to bool
    - media: i2c: ov2311: Fix uninitialized variable usage
    - drm/panel: Fix default values for Waveshare 7.9 inch DSI touchscreen (#5565)
    - dtoverlays: Add i2c bus overrides to edt-ft5406 overlay
    - dtoverlays: Fix README text for i2c-fan
    - drivers: irqchip: irq-bcm2835: Concurrency fix
    - defconfigs: Add TINYDRM_ILI9486 to defconfigs
    - dtoverlays: Add drm option to piscreen overlay
    - drm/ili9486: Resolve clash in spi_device_id names
    - input: ads7846: Add missing spi_device_id strings
    - staging: bcm2835-codec: Downgrade the level for a debug message
    - configs: Raise 8250 UART limit to 5 on BCM2711

  [ Ubuntu: 6.5.0-5.5 ]

  * mantic/linux: 6.5.0-5.5 -proposed tracker (LP: #2034546)
  * Packaging resync (LP: #1786013)
    - [Packaging] update helper scripts
    - debian/dkms-versions -- update from kernel-versions (main/d2023.08.23)

  [ Ubuntu: 6.5.0-4.4 ]

  * mantic/linux: 6.5.0-4.4 -proposed tracker (LP: #2034042)
  * Packaging resync (LP: #1786013)
    - debian/dkms-versions -- update from kernel-versions (main/d2023.08.23)

  [ Ubuntu: 6.5.0-3.3 ]

  * mantic/linux: 6.5.0-3.3 -proposed tracker (LP: #2033904)
  * Packaging resync (LP: #1786013)
    - debian/dkms-versions -- update from kernel-versions (main/d2023.08.23)
  * [23.10] Please test secure-boot and lockdown on the early 6.5 kernel (s390x)
    (LP: #2026833)
    - [Packaging] re-enable signing for s390x
  * Miscellaneous upstream changes
    - module/decompress: use vmalloc() for zstd decompression workspace

  [ Ubuntu: 6.5.0-2.2 ]

  * mantic/linux: 6.5.0-2.2 -proposed tracker (LP: #2033240)
  * Soundwire support for Dell SKU0C87 devices (LP: #2029281)
    - SAUCE: ASoC: Intel: soc-acpi: add support for Dell SKU0C87 devices
  * Fix numerous AER related issues (LP: #2033025)
    - SAUCE: PCI/AER: Disable AER service during suspend, again
    - SAUCE: PCI/DPC: Disable DPC service during suspend, again
  * Support Realtek RTL8852CE WiFi 6E/BT Combo (LP: #2025672)
    - wifi: r...

Changed in linux-raspi (Ubuntu Mantic):
status: New → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.