md: improve IO accounting

This bug is missing log files that will aid in diagnosing the problem. While running an Ubuntu kernel (not a mainline or third-party kernel) please enter the following command in a terminal window:

apport-collect 1891151

and then change the status of the bug to 'Confirmed'.

If, due to the nature of the issue you have encountered, you are unable to run this command, please add a comment stating that fact and change the bug status to 'Confirmed'.

This change has been made by an automated script, maintained by the Ubuntu Kernel Team.

Not cleanly cherry picked back to 5.8 or 5.4, so I'm going to push this to the kernel team to investigate.

This bug is missing log files that will aid in diagnosing the problem. While running an Ubuntu kernel (not a mainline or third-party kernel) please enter the following command in a terminal window:

apport-collect 1891151

and then change the status of the bug to 'Confirmed'.

If, due to the nature of the issue you have encountered, you are unable to run this command, please add a comment stating that fact and change the bug status to 'Confirmed'.

This change has been made by an automated script, maintained by the Ubuntu Kernel Team.

For 5.8 this is a simple backport with one conflict caused by function md_make_request() renamed to md_submit_bio(). Then it can be easily cherry picked to 5.4 but will need a change to use generic_{start,end}_io_acct() instead of disk_{start,end}_io_acct(). I can send you the backported patches if you wish.

@mtkaczyk,

Yes, it would expedite the process if you could provide the patches. Thanks for offering.

The Groovy version is already on the ml, looking at the Focal version now but s/disk_{start,end}_io_acct/generic_{start,end}_io_acct/g is not enough.

FYI, the patch on the mailing list is here: https://lists.ubuntu.com/archives/kernel-team/2020-September/113590.html

This bug was fixed in the package linux - 5.8.0-20.21

---------------
linux (5.8.0-20.21) groovy; urgency=medium

  * groovy/linux: 5.8.0-20.21 -proposed tracker (LP: #1896668)

  * Lenovo ThinkBook 14-IML Touchpad not showing up in /proc/bus/input/devices
    (LP: #1853277)
    - i2c: core: Call i2c_acpi_install_space_handler() before
      i2c_acpi_register_devices()

  * Enable LTR for endpoints behind VMD (LP: #1896598)
    - SAUCE: PCI/ASPM: Enable LTR for endpoints behind VMD

  * Remove duplicated code in ip_defrag.sh of kselftests/net (LP: #1894062)
    - Revert "UBUNTU: SAUCE: selftests: net: ip_defrag: modprobe missing
      nf_defrag_ipv6 support"

  * [SRU] [Focal/OEM-5.6/Groovy]Fix AMD usb host controller lost after stress S3
    (LP: #1893914)
    - SAUCE: xhci: workaround for S3 issue on AMD SNPS 3.0 xHC

  * debian/rules editconfigs does not work on s390x to change s390x only configs
    (LP: #1863116)
    - [Packaging] kernelconfig -- only update/edit configurations on architectures
      we have compiler support

  * [Ubuntu 20.10] zPCI DMA tables and bitmap leak on hard unplug (PCI Event
    0x0304) (LP: #1896216)
    - s390/pci: fix leak of DMA tables on hard unplug

  * md: improve IO accounting (LP: #1891151)
    - md: improve io stats accounting

  * Groovy update: v5.8.10 upstream stable release (LP: #1896078)
    - ARM: OMAP2+: Fix an IS_ERR() vs NULL check in _get_pwrdm()
    - ARM: dts: logicpd-torpedo-baseboard: Fix broken audio
    - ARM: dts: logicpd-som-lv-baseboard: Fix broken audio
    - ARM: dts: logicpd-som-lv-baseboard: Fix missing video
    - regulator: push allocation in regulator_ena_gpio_request() out of lock
    - regulator: remove superfluous lock in regulator_resolve_coupling()
    - ARM: dts: socfpga: fix register entry for timer3 on Arria10
    - ARM: dts: omap5: Fix DSI base address and clocks
    - ARM: dts: ls1021a: fix QuadSPI-memory reg range
    - ARM: dts: imx7ulp: Correct gpio ranges
    - arm64: dts: imx: Add missing imx8mm-beacon-kit.dtb to build
    - ARM: dts: imx7d-zii-rmu2: fix rgmii phy-mode for ksz9031 phy
    - RDMA/rtrs-srv: Replace device_register with device_initialize and device_add
    - RDMA/rxe: Fix memleak in rxe_mem_init_user
    - RDMA/rxe: Drop pointless checks in rxe_init_ports
    - RDMA/rxe: Fix panic when calling kmem_cache_create()
    - RDMA/bnxt_re: Do not report transparent vlan from QP1
    - RDMA/bnxt_re: Fix the qp table indexing
    - RDMA/bnxt_re: Static NQ depth allocation
    - RDMA/bnxt_re: Fix driver crash on unaligned PSN entry address
    - RDMA/bnxt_re: Remove the qp from list only if the qp destroy succeeds
    - drm/sun4i: add missing put_device() call in sun8i_r40_tcon_tv_set_mux()
    - arm64: dts: imx8mq: Fix TMU interrupt property
    - drm/sun4i: Fix dsi dcs long write function
    - scsi: qla2xxx: Fix regression on sparc64
    - scsi: libsas: Set data_dir as DMA_NONE if libata marks qc as NODATA
    - drm/virtio: fix unblank
    - RDMA/core: Fix unsafe linked list traversal after failing to allocate CQ
    - RDMA/core: Fix reported speed and width
    - scsi: megaraid_sas: Don't call disable_irq from process IRQ poll
    - scsi: mpt3sas: Don'...