Excessive size of kernel modules on RISC-V - modules unstripped

[Heinrich Schuchardt]

The size of kernel modules on RISC-V indicates that there is a build problem:

kernel/drivers/ata/sata_mv.ko
amd64: 81713 bytes
arm64: 85505 bytes
riscv64: 601417 bytes

objdump -h shows that there is no section beyond 0xc6fc in the riscv64 file.
The size of the sections matches what we see on the other architectures.

[Dimitri John Ledkov (xnox)]

Modules appear to be unstripped. Debugging why that it is so:

$ file ./lib/modules/5.15.0-1004-generic/kernel/fs/btrfs/btrfs.ko
./lib/modules/5.15.0-1004-generic/kernel/fs/btrfs/btrfs.ko: ELF 64-bit LSB relocatable, UCB RISC-V, version 1 (SYSV), BuildID[sha1]=6c48292f27241752d4993a6854e9eee38ab3b503, not stripped

$ du ./lib/modules/5.15.0-1004-generic/kernel/fs/btrfs/btrfs.ko
16968 ./lib/modules/5.15.0-1004-generic/kernel/fs/btrfs/btrfs.ko

$ riscv64-linux-gnu-strip ./lib/modules/5.15.0-1004-generic/kernel/fs/btrfs/btrfs.ko

$ file ./lib/modules/5.15.0-1004-generic/kernel/fs/btrfs/btrfs.ko
./lib/modules/5.15.0-1004-generic/kernel/fs/btrfs/btrfs.ko: ELF 64-bit LSB relocatable, UCB RISC-V, version 1 (SYSV), BuildID[sha1]=6c48292f27241752d4993a6854e9eee38ab3b503, stripped

$ du ./lib/modules/5.15.0-1004-generic/kernel/fs/btrfs/btrfs.ko
1384 ./lib/modules/5.15.0-1004-generic/kernel/fs/btrfs/btrfs.ko

[Heinrich Schuchardt (xypron)]

If I compile Linux v5.15.26 with RISC-V defconfig + CONFIG_SATA_MV=y,
I get a drivers/ata/sata_mv.ko with 103960 bytes.

With additionally CONFIG_MODULE_SIG=y after make modules_install: 104673 bytes.

So the problems seems not to exist in upstream.

[Dimitri John Ledkov (xnox)]

we are investing things further.
it seems like debug symbol stripping is fine.
but the input .ko are huge, and much larger than on other arches, for currently unknown reason.

[Heinrich Schuchardt (xypron)]

Cf. [PATCH] UBUNTU: SAUCE: scripts/Makefile.modinst discard-locals from modules
 (https://lists.ubuntu.com/archives/kernel-team/2022-March/128534.html)

[Heinrich Schuchardt (xypron)]

Cf. "Why is .symtab so huge on riscv?" #1036 (https://github.com/riscv-collab/riscv-gnu-toolchain/issues/1036)

[Launchpad Janitor (janitor)]

This bug was fixed in the package linux-riscv - 5.15.0-1005.5

---------------
linux-riscv (5.15.0-1005.5) jammy; urgency=medium

  * jammy/linux-riscv: 5.15.0-1005.5 -proposed tracker (LP: #1965783)

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

  * dependency on crda obsolete according to Debian (LP: #1958918)
    - [Packaging] switch dependency from crda to wireless-regdb

  * Excessive size of kernel modules on RISC-V - modules unstripped
    (LP: #1964335)
    - SAUCE: scripts/Makefile.modinst discard-locals from modules

  * Cherrypick features and fixes from v5.17 for sifive dtbs (LP: #1962010)
    - riscv: dts: sifive: use only generic JEDEC SPI NOR flash compatible
    - riscv: dts: sifive: fix Unleashed board compatible
    - Revert "SiFive HiFive Unleashed: Add PWM LEDs (D1, D2, D3, D4)"
    - riscv: dts: sifive: drop duplicated nodes and properties in sifive
    - riscv: dts: sifive: add missing compatible for plic
    - riscv: dts: sifive unmatched: Name gpio lines
    - riscv: dts: sifive unmatched: Expose the board ID eeprom
    - riscv: dts: sifive unmatched: Expose the PMIC sub-functions
    - Revert "riscv: sifive: unmatched: update for 16GB rev3"
    - riscv: dts: sifive unmatched: Fix regulator for board rev3
    - riscv: dts: sifive unmatched: Link the tmp451 with its power supply
    - riscv: dts: sifive: Group tuples in interrupt properties
    - riscv: dts: sifive: Group tuples in register properties
    - riscv: dts: sifive: fu540-c000: Drop bogus soc node compatible values
    - riscv: dts: sifive: fu540-c000: Fix PLIC node
    - riscv: dts: sifive unmatched: Add gpio poweroff

  * riscv: backport support for SiFive Unmatched (LP: #1910965)
    - SiFive HiFive Unleashed: Add PWM LEDs (D1, D2, D3, D4)

  * Miscellaneous Ubuntu changes
    - [Config] update configs and annotations after rebase

  * Miscellaneous upstream changes
    - RISC-V: Use SBI SRST extension when available

  [ Ubuntu: 5.15.0-23.23 ]

  * jammy/linux: 5.15.0-23.23 -proposed tracker (LP: #1964573)
  * Packaging resync (LP: #1786013)
    - [Packaging] resync dkms-build{,--nvidia-N} from LRMv5
    - debian/dkms-versions -- update from kernel-versions (main/master)
  * [22.04 FEAT] KVM: Enable GISA support for Secure Execution guests
    (LP: #1959977)
    - KVM: s390: pv: make use of ultravisor AIV support
  * intel_iommu breaks Intel IPU6 camera: isys port open ready failed -16
    (LP: #1958004)
    - SAUCE: iommu: intel-ipu: use IOMMU passthrough mode for Intel IPUs
  * CVE-2022-23960
    - ARM: report Spectre v2 status through sysfs
    - ARM: early traps initialisation
    - ARM: use LOADADDR() to get load address of sections
    - ARM: Spectre-BHB workaround
    - ARM: include unprivileged BPF status in Spectre V2 reporting
    - arm64: Add Neoverse-N2, Cortex-A710 CPU part definition
    - arm64: Add HWCAP for self-synchronising virtual counter
    - arm64: Add Cortex-X2 CPU part definition
    - arm64: add ID_AA64ISAR2_EL1 sys register
    - arm64: cpufeature: add HWCAP for FEAT_AFP
    - arm64: cpufeature: add HWCAP for FEAT_RPRES
    - arm64: entry.S: Add ventry overflow sanity checks
    - arm64: spectre: Rename s...

[Ubuntu Kernel Bot (ubuntu-kernel-bot)]

This bug is awaiting verification that the linux-nezha-5.17/5.17.0-1001.1 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!

[Dimitri John Ledkov (xnox)]

verified that linux-nezha-5.17/5.17.0-1001.1 has small .ko modules.

[Dimitri John Ledkov (xnox)]

and allwinner; and starfive.

[Ubuntu Kernel Bot (ubuntu-kernel-bot)]

This bug is awaiting verification that the linux-starfive/6.2.0-1002.2 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!