diff mbox series

[v6,02/10] kbuild: link symbol CRCs at final link, removing CONFIG_MODULE_REL_CRCS

Message ID 20220513113930.10488-3-masahiroy@kernel.org (mailing list archive)
State New, archived
Headers show
Series kbuild: yet another series of cleanups (modpost, LTO, MODULE_REL_CRCS, export.h) | expand

Commit Message

Masahiro Yamada May 13, 2022, 11:39 a.m. UTC
include/{linux,asm-generic}/export.h defines a weak symbol, __crc_*
as a placeholder.

Genksyms writes the version CRCs into the linker script, which will be
used for filling the __crc_* symbols. The linker script format depends
on CONFIG_MODULE_REL_CRCS. If it is enabled, __crc_* holds the offset
to the reference of CRC.

It is time to get rid of this complexity.

Now that modpost parses text files (.*.cmd) to collect all the CRCs,
it can generate C code that will be linked to the vmlinux or modules.

Generate a new C file, .vmlinux.export.c, which contains the CRCs of
symbols exported by vmlinux. It is compiled and linked to vmlinux in
scripts/link-vmlinux.sh.

Put the CRCs of symbols exported by modules into the existing *.mod.c
files. No additional build step is needed for modules. As before,
*.mod.c are compiled and linked to *.ko in scripts/Makefile.modfinal.

No linker magic is used here. The new C implementation works in the
same way, whether CONFIG_RELOCATABLE is enabled or not.
CONFIG_MODULE_REL_CRCS is no longer needed.

Previously, Kbuild invoked additional $(LD) to update the CRCs in
objects, but this step is unneeded too.

Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Tested-by: Nathan Chancellor <nathan@kernel.org>
Tested-by: Nicolas Schier <nicolas@fjasle.eu>
Reviewed-by: Nicolas Schier <nicolas@fjasle.eu>
---

(no changes since v5)

Changes in v5:
  - Fix the build error when CONFIG_DEBUG_INFO_BTF=y (reported by Nathan)
  - Clean up arch/m68k/include/asm/export.h (Nick)
  - Keep gen_symversions (and will be removed by a later patch)

Changes in v4:
  - Rename .vmlinux-symver.c to .vmlinux.export.c
    because I notice this approach is useful for further cleanups,
    not only for modversioning but also for overall EXPORT_SYMBOL.

Changes in v3:
  - New patch

 arch/m68k/include/asm/Kbuild    |  1 +
 arch/m68k/include/asm/export.h  |  2 --
 arch/powerpc/Kconfig            |  1 -
 arch/s390/Kconfig               |  1 -
 arch/um/Kconfig                 |  1 -
 include/asm-generic/export.h    | 22 ++++++++--------------
 include/linux/export-internal.h | 16 ++++++++++++++++
 include/linux/export.h          | 30 ++++++++----------------------
 init/Kconfig                    |  4 ----
 kernel/module.c                 | 10 +---------
 scripts/Makefile.build          | 27 ++++-----------------------
 scripts/genksyms/genksyms.c     | 18 ++++--------------
 scripts/link-vmlinux.sh         | 14 +++++++++++++-
 scripts/mod/modpost.c           | 28 ++++++++++++++++++++++++----
 14 files changed, 79 insertions(+), 96 deletions(-)
 delete mode 100644 arch/m68k/include/asm/export.h
 create mode 100644 include/linux/export-internal.h

Comments

Christophe Leroy Aug. 20, 2022, 10:02 a.m. UTC | #1
Hi,

Le 13/05/2022 à 13:39, Masahiro Yamada a écrit :
> include/{linux,asm-generic}/export.h defines a weak symbol, __crc_*
> as a placeholder.
> 
> Genksyms writes the version CRCs into the linker script, which will be
> used for filling the __crc_* symbols. The linker script format depends
> on CONFIG_MODULE_REL_CRCS. If it is enabled, __crc_* holds the offset
> to the reference of CRC.
> 
> It is time to get rid of this complexity.
> 
> Now that modpost parses text files (.*.cmd) to collect all the CRCs,
> it can generate C code that will be linked to the vmlinux or modules.
> 
> Generate a new C file, .vmlinux.export.c, which contains the CRCs of
> symbols exported by vmlinux. It is compiled and linked to vmlinux in
> scripts/link-vmlinux.sh.
> 
> Put the CRCs of symbols exported by modules into the existing *.mod.c
> files. No additional build step is needed for modules. As before,
> *.mod.c are compiled and linked to *.ko in scripts/Makefile.modfinal.
> 
> No linker magic is used here. The new C implementation works in the
> same way, whether CONFIG_RELOCATABLE is enabled or not.
> CONFIG_MODULE_REL_CRCS is no longer needed.
> 
> Previously, Kbuild invoked additional $(LD) to update the CRCs in
> objects, but this step is unneeded too.
> 
> Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
> Tested-by: Nathan Chancellor <nathan@kernel.org>
> Tested-by: Nicolas Schier <nicolas@fjasle.eu>
> Reviewed-by: Nicolas Schier <nicolas@fjasle.eu>

Problem with v6.0-rc1
Problem with v5.19
No problem with v5.18

Bisected to 7b4537199a4a ("kbuild: link symbol CRCs at final link, 
removing CONFIG_MODULE_REL_CRCS")

The above patch leads to the following problem building 
mpc85xx_defconfig + CONFIG_RELOCATABLE

   LD      vmlinux
   SYSMAP  System.map
   SORTTAB vmlinux
   CHKREL  vmlinux
WARNING: 451 bad relocations
c0b0f26d R_PPC_UADDR32     .head.text-0x3ff9f2bc
c0b0f271 R_PPC_UADDR32     .head.text-0x3ffac300
c0b0f275 R_PPC_UADDR32     .head.text-0x3ffb0bdc
c0b0f279 R_PPC_UADDR32     .head.text-0x3fe1e080
c0b0f27d R_PPC_UADDR32     .head.text-0x3fe1df4c
c0b0f281 R_PPC_UADDR32     .head.text-0x3fe21514
c0b0f285 R_PPC_UADDR32     .head.text-0x3fe211c0
c0b0f289 R_PPC_UADDR32     .head.text-0x3ffabda0
c0b0f28d R_PPC_UADDR32     .head.text-0x3fe21258
c0b0f291 R_PPC_UADDR32     .head.text-0x3fe074d0
c0b0f295 R_PPC_UADDR32     .head.text-0x3fe07ad4
c0b0f299 R_PPC_UADDR32     .head.text-0x3fe13470
c0b0f29d R_PPC_UADDR32     .head.text-0x3fe22700
c0b0f2a1 R_PPC_UADDR32     .head.text-0x3ff4b8e0
c0b0f2a5 R_PPC_UADDR32     .head.text-0x3fe08320
c0b0f2a9 R_PPC_UADDR32     .head.text-0x3fe220dc
c0b0f2ad R_PPC_UADDR32     .head.text-0x3fe21da0
c0b0f2b1 R_PPC_UADDR32     .head.text-0x3ff89dc0
c0b0f2b5 R_PPC_UADDR32     .head.text-0x3fe16524
c0b0f2b9 R_PPC_UADDR32     .head.text-0x3fe1ef74
c0b0f2bd R_PPC_UADDR32     .head.text-0x3ff98b84
c0b0f2c1 R_PPC_UADDR32     .head.text-0x3fdef9a0
c0b0f2c5 R_PPC_UADDR32     .head.text-0x3fdf21ac
c0b0f2c9 R_PPC_UADDR32     .head.text-0x3ff993c4
...
c0b0f969 R_PPC_UADDR32     .head.text-0x3ff89dc0
c0b0f96d R_PPC_UADDR32     .head.text-0x3fe9ad40
c0b0f971 R_PPC_UADDR32     .head.text-0x3ff2eb00
c0b0f975 R_PPC_UADDR32     .head.text-0x3ff89dc0

And boot fails:

Run /init as init process
kernel tried to execute user page (0) - exploit attempt? (uid: 0)
BUG: Unable to handle kernel instruction fetch (NULL pointer?)
Faulting instruction address: 0x00000000
Oops: Kernel access of bad area, sig: 11 [#1]
BE PAGE_SIZE=4K MPC8544 DS
Modules linked in:
CPU: 0 PID: 1 Comm: init Not tainted 5.18.0-rc1-00054-g7b4537199a4a #1523
NIP:  00000000 LR: c00150e4 CTR: 00000000
REGS: c3091e10 TRAP: 0400   Not tainted  (5.18.0-rc1-00054-g7b4537199a4a)
MSR:  00009000 <EE,ME>  CR: 88000422  XER: 20000000

GPR00: 00004000 c3091f00 c30c8000 00000000 00000013 b7bb9f4c b7bd8f60 
bfee6650
GPR08: 00000054 00000000 c0b0f26d 00000000 c13b0000 00000000 bfee6668 
00000000
GPR16: 84e08000 00000000 08000000 00000064 00000000 00102000 00000001 
00000001
GPR24: 00000001 00000001 b7b9c7d0 10000034 00000009 b7bd8f38 b7bd9854 
b7bd8688
NIP [00000000] 0x0
LR [c00150e4] ret_from_syscall+0x0/0x28
Call Trace:
[c3091f00] [c0000af0] InstructionStorage+0x150/0x160 (unreliable)
--- interrupt: c00 at 0xb7bb28e8
NIP:  b7bb28e8 LR: b7bb1384 CTR: b7bb1218
REGS: c3091f10 TRAP: 0c00   Not tainted  (5.18.0-rc1-00054-g7b4537199a4a)
MSR:  0002d000 <CE,EE,PR,ME>  CR: 28000422  XER: 20000000

GPR00: 0000002d bfee61f0 00000000 00000000 00000013 b7bb9f4c b7bd8f60 
bfee6650
GPR08: 00000054 00000020 bfee6648 00000000 00000001 00000000 bfee6668 
00000000
GPR16: 84e08000 00000000 08000000 00000064 00000000 00102000 00000001 
00000001
GPR24: 00000001 00000001 b7b9c7d0 10000034 00000009 b7bd8f38 b7bd9854 
b7bd8688
NIP [b7bb28e8] 0xb7bb28e8
LR [b7bb1384] 0xb7bb1384
--- interrupt: c00
Instruction dump:
XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX
XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX
---[ end trace 0000000000000000 ]---

Kernel panic - not syncing: Attempted to kill init! exitcode=0x0000000b



Christophe
Sedat Dilek Aug. 20, 2022, 12:05 p.m. UTC | #2
On Sat, Aug 20, 2022 at 12:04 PM Christophe Leroy
<christophe.leroy@csgroup.eu> wrote:
>
> Hi,
>
> Le 13/05/2022 à 13:39, Masahiro Yamada a écrit :
> > include/{linux,asm-generic}/export.h defines a weak symbol, __crc_*
> > as a placeholder.
> >
> > Genksyms writes the version CRCs into the linker script, which will be
> > used for filling the __crc_* symbols. The linker script format depends
> > on CONFIG_MODULE_REL_CRCS. If it is enabled, __crc_* holds the offset
> > to the reference of CRC.
> >
> > It is time to get rid of this complexity.
> >
> > Now that modpost parses text files (.*.cmd) to collect all the CRCs,
> > it can generate C code that will be linked to the vmlinux or modules.
> >
> > Generate a new C file, .vmlinux.export.c, which contains the CRCs of
> > symbols exported by vmlinux. It is compiled and linked to vmlinux in
> > scripts/link-vmlinux.sh.
> >
> > Put the CRCs of symbols exported by modules into the existing *.mod.c
> > files. No additional build step is needed for modules. As before,
> > *.mod.c are compiled and linked to *.ko in scripts/Makefile.modfinal.
> >
> > No linker magic is used here. The new C implementation works in the
> > same way, whether CONFIG_RELOCATABLE is enabled or not.
> > CONFIG_MODULE_REL_CRCS is no longer needed.
> >
> > Previously, Kbuild invoked additional $(LD) to update the CRCs in
> > objects, but this step is unneeded too.
> >
> > Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
> > Tested-by: Nathan Chancellor <nathan@kernel.org>
> > Tested-by: Nicolas Schier <nicolas@fjasle.eu>
> > Reviewed-by: Nicolas Schier <nicolas@fjasle.eu>
>
> Problem with v6.0-rc1
> Problem with v5.19
> No problem with v5.18
>
> Bisected to 7b4537199a4a ("kbuild: link symbol CRCs at final link,
> removing CONFIG_MODULE_REL_CRCS")
>

What you are looking for is...

commit 7d13fd96df875a9d786ee6dcc8fec460d35d4b12
("modpost: fix module versioning when a symbol lacks valid CRC")

It's pending in kbuild.git#fixes.

-Sedat-

[1] https://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild.git/commit/?h=fixes&id=7d13fd96df875a9d786ee6dcc8fec460d35d4b12
[2] https://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild.git/log/?h=fixes

> The above patch leads to the following problem building
> mpc85xx_defconfig + CONFIG_RELOCATABLE
>
>    LD      vmlinux
>    SYSMAP  System.map
>    SORTTAB vmlinux
>    CHKREL  vmlinux
> WARNING: 451 bad relocations
> c0b0f26d R_PPC_UADDR32     .head.text-0x3ff9f2bc
> c0b0f271 R_PPC_UADDR32     .head.text-0x3ffac300
> c0b0f275 R_PPC_UADDR32     .head.text-0x3ffb0bdc
> c0b0f279 R_PPC_UADDR32     .head.text-0x3fe1e080
> c0b0f27d R_PPC_UADDR32     .head.text-0x3fe1df4c
> c0b0f281 R_PPC_UADDR32     .head.text-0x3fe21514
> c0b0f285 R_PPC_UADDR32     .head.text-0x3fe211c0
> c0b0f289 R_PPC_UADDR32     .head.text-0x3ffabda0
> c0b0f28d R_PPC_UADDR32     .head.text-0x3fe21258
> c0b0f291 R_PPC_UADDR32     .head.text-0x3fe074d0
> c0b0f295 R_PPC_UADDR32     .head.text-0x3fe07ad4
> c0b0f299 R_PPC_UADDR32     .head.text-0x3fe13470
> c0b0f29d R_PPC_UADDR32     .head.text-0x3fe22700
> c0b0f2a1 R_PPC_UADDR32     .head.text-0x3ff4b8e0
> c0b0f2a5 R_PPC_UADDR32     .head.text-0x3fe08320
> c0b0f2a9 R_PPC_UADDR32     .head.text-0x3fe220dc
> c0b0f2ad R_PPC_UADDR32     .head.text-0x3fe21da0
> c0b0f2b1 R_PPC_UADDR32     .head.text-0x3ff89dc0
> c0b0f2b5 R_PPC_UADDR32     .head.text-0x3fe16524
> c0b0f2b9 R_PPC_UADDR32     .head.text-0x3fe1ef74
> c0b0f2bd R_PPC_UADDR32     .head.text-0x3ff98b84
> c0b0f2c1 R_PPC_UADDR32     .head.text-0x3fdef9a0
> c0b0f2c5 R_PPC_UADDR32     .head.text-0x3fdf21ac
> c0b0f2c9 R_PPC_UADDR32     .head.text-0x3ff993c4
> ...
> c0b0f969 R_PPC_UADDR32     .head.text-0x3ff89dc0
> c0b0f96d R_PPC_UADDR32     .head.text-0x3fe9ad40
> c0b0f971 R_PPC_UADDR32     .head.text-0x3ff2eb00
> c0b0f975 R_PPC_UADDR32     .head.text-0x3ff89dc0
>
> And boot fails:
>
> Run /init as init process
> kernel tried to execute user page (0) - exploit attempt? (uid: 0)
> BUG: Unable to handle kernel instruction fetch (NULL pointer?)
> Faulting instruction address: 0x00000000
> Oops: Kernel access of bad area, sig: 11 [#1]
> BE PAGE_SIZE=4K MPC8544 DS
> Modules linked in:
> CPU: 0 PID: 1 Comm: init Not tainted 5.18.0-rc1-00054-g7b4537199a4a #1523
> NIP:  00000000 LR: c00150e4 CTR: 00000000
> REGS: c3091e10 TRAP: 0400   Not tainted  (5.18.0-rc1-00054-g7b4537199a4a)
> MSR:  00009000 <EE,ME>  CR: 88000422  XER: 20000000
>
> GPR00: 00004000 c3091f00 c30c8000 00000000 00000013 b7bb9f4c b7bd8f60
> bfee6650
> GPR08: 00000054 00000000 c0b0f26d 00000000 c13b0000 00000000 bfee6668
> 00000000
> GPR16: 84e08000 00000000 08000000 00000064 00000000 00102000 00000001
> 00000001
> GPR24: 00000001 00000001 b7b9c7d0 10000034 00000009 b7bd8f38 b7bd9854
> b7bd8688
> NIP [00000000] 0x0
> LR [c00150e4] ret_from_syscall+0x0/0x28
> Call Trace:
> [c3091f00] [c0000af0] InstructionStorage+0x150/0x160 (unreliable)
> --- interrupt: c00 at 0xb7bb28e8
> NIP:  b7bb28e8 LR: b7bb1384 CTR: b7bb1218
> REGS: c3091f10 TRAP: 0c00   Not tainted  (5.18.0-rc1-00054-g7b4537199a4a)
> MSR:  0002d000 <CE,EE,PR,ME>  CR: 28000422  XER: 20000000
>
> GPR00: 0000002d bfee61f0 00000000 00000000 00000013 b7bb9f4c b7bd8f60
> bfee6650
> GPR08: 00000054 00000020 bfee6648 00000000 00000001 00000000 bfee6668
> 00000000
> GPR16: 84e08000 00000000 08000000 00000064 00000000 00102000 00000001
> 00000001
> GPR24: 00000001 00000001 b7b9c7d0 10000034 00000009 b7bd8f38 b7bd9854
> b7bd8688
> NIP [b7bb28e8] 0xb7bb28e8
> LR [b7bb1384] 0xb7bb1384
> --- interrupt: c00
> Instruction dump:
> XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX
> XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX
> ---[ end trace 0000000000000000 ]---
>
> Kernel panic - not syncing: Attempted to kill init! exitcode=0x0000000b
>
>
>
> Christophe
Masahiro Yamada Aug. 20, 2022, 12:51 p.m. UTC | #3
On Sat, Aug 20, 2022 at 7:02 PM Christophe Leroy
<christophe.leroy@csgroup.eu> wrote:
>
> Hi,
>
> Le 13/05/2022 à 13:39, Masahiro Yamada a écrit :
> > include/{linux,asm-generic}/export.h defines a weak symbol, __crc_*
> > as a placeholder.
> >
> > Genksyms writes the version CRCs into the linker script, which will be
> > used for filling the __crc_* symbols. The linker script format depends
> > on CONFIG_MODULE_REL_CRCS. If it is enabled, __crc_* holds the offset
> > to the reference of CRC.
> >
> > It is time to get rid of this complexity.
> >
> > Now that modpost parses text files (.*.cmd) to collect all the CRCs,
> > it can generate C code that will be linked to the vmlinux or modules.
> >
> > Generate a new C file, .vmlinux.export.c, which contains the CRCs of
> > symbols exported by vmlinux. It is compiled and linked to vmlinux in
> > scripts/link-vmlinux.sh.
> >
> > Put the CRCs of symbols exported by modules into the existing *.mod.c
> > files. No additional build step is needed for modules. As before,
> > *.mod.c are compiled and linked to *.ko in scripts/Makefile.modfinal.
> >
> > No linker magic is used here. The new C implementation works in the
> > same way, whether CONFIG_RELOCATABLE is enabled or not.
> > CONFIG_MODULE_REL_CRCS is no longer needed.
> >
> > Previously, Kbuild invoked additional $(LD) to update the CRCs in
> > objects, but this step is unneeded too.
> >
> > Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
> > Tested-by: Nathan Chancellor <nathan@kernel.org>
> > Tested-by: Nicolas Schier <nicolas@fjasle.eu>
> > Reviewed-by: Nicolas Schier <nicolas@fjasle.eu>
>
> Problem with v6.0-rc1
> Problem with v5.19
> No problem with v5.18
>
> Bisected to 7b4537199a4a ("kbuild: link symbol CRCs at final link,
> removing CONFIG_MODULE_REL_CRCS")
>
> The above patch leads to the following problem building
> mpc85xx_defconfig + CONFIG_RELOCATABLE



Is this because the relocation implementation on ppc is incomplete?
(and is it the reason why relock_check.sh exists?)

arch/powerpc/kernel/reloc_32.S does not support R_PPC_UADDR32
Christophe Leroy Aug. 20, 2022, 2:10 p.m. UTC | #4
Le 20/08/2022 à 14:05, Sedat Dilek a écrit :
> On Sat, Aug 20, 2022 at 12:04 PM Christophe Leroy
> <christophe.leroy@csgroup.eu> wrote:
>>
>> Hi,
>>
>> Le 13/05/2022 à 13:39, Masahiro Yamada a écrit :
>>> include/{linux,asm-generic}/export.h defines a weak symbol, __crc_*
>>> as a placeholder.
>>>
>>> Genksyms writes the version CRCs into the linker script, which will be
>>> used for filling the __crc_* symbols. The linker script format depends
>>> on CONFIG_MODULE_REL_CRCS. If it is enabled, __crc_* holds the offset
>>> to the reference of CRC.
>>>
>>> It is time to get rid of this complexity.
>>>
>>> Now that modpost parses text files (.*.cmd) to collect all the CRCs,
>>> it can generate C code that will be linked to the vmlinux or modules.
>>>
>>> Generate a new C file, .vmlinux.export.c, which contains the CRCs of
>>> symbols exported by vmlinux. It is compiled and linked to vmlinux in
>>> scripts/link-vmlinux.sh.
>>>
>>> Put the CRCs of symbols exported by modules into the existing *.mod.c
>>> files. No additional build step is needed for modules. As before,
>>> *.mod.c are compiled and linked to *.ko in scripts/Makefile.modfinal.
>>>
>>> No linker magic is used here. The new C implementation works in the
>>> same way, whether CONFIG_RELOCATABLE is enabled or not.
>>> CONFIG_MODULE_REL_CRCS is no longer needed.
>>>
>>> Previously, Kbuild invoked additional $(LD) to update the CRCs in
>>> objects, but this step is unneeded too.
>>>
>>> Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
>>> Tested-by: Nathan Chancellor <nathan@kernel.org>
>>> Tested-by: Nicolas Schier <nicolas@fjasle.eu>
>>> Reviewed-by: Nicolas Schier <nicolas@fjasle.eu>
>>
>> Problem with v6.0-rc1
>> Problem with v5.19
>> No problem with v5.18
>>
>> Bisected to 7b4537199a4a ("kbuild: link symbol CRCs at final link,
>> removing CONFIG_MODULE_REL_CRCS")
>>
> 
> What you are looking for is...
> 
> commit 7d13fd96df875a9d786ee6dcc8fec460d35d4b12
> ("modpost: fix module versioning when a symbol lacks valid CRC")
> 
> It's pending in kbuild.git#fixes.
> 
> -Sedat-
> 
> [1] https://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild.git/commit/?h=fixes&id=7d13fd96df875a9d786ee6dcc8fec460d35d4b12
> [2] https://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild.git/log/?h=fixes
> 

That patch doesn't fix the problem.

Christophe
Christophe Leroy Aug. 20, 2022, 2:15 p.m. UTC | #5
Le 20/08/2022 à 14:51, Masahiro Yamada a écrit :
> On Sat, Aug 20, 2022 at 7:02 PM Christophe Leroy
> <christophe.leroy@csgroup.eu> wrote:
>>
>> Hi,
>>
>> Le 13/05/2022 à 13:39, Masahiro Yamada a écrit :
>>> include/{linux,asm-generic}/export.h defines a weak symbol, __crc_*
>>> as a placeholder.
>>>
>>> Genksyms writes the version CRCs into the linker script, which will be
>>> used for filling the __crc_* symbols. The linker script format depends
>>> on CONFIG_MODULE_REL_CRCS. If it is enabled, __crc_* holds the offset
>>> to the reference of CRC.
>>>
>>> It is time to get rid of this complexity.
>>>
>>> Now that modpost parses text files (.*.cmd) to collect all the CRCs,
>>> it can generate C code that will be linked to the vmlinux or modules.
>>>
>>> Generate a new C file, .vmlinux.export.c, which contains the CRCs of
>>> symbols exported by vmlinux. It is compiled and linked to vmlinux in
>>> scripts/link-vmlinux.sh.
>>>
>>> Put the CRCs of symbols exported by modules into the existing *.mod.c
>>> files. No additional build step is needed for modules. As before,
>>> *.mod.c are compiled and linked to *.ko in scripts/Makefile.modfinal.
>>>
>>> No linker magic is used here. The new C implementation works in the
>>> same way, whether CONFIG_RELOCATABLE is enabled or not.
>>> CONFIG_MODULE_REL_CRCS is no longer needed.
>>>
>>> Previously, Kbuild invoked additional $(LD) to update the CRCs in
>>> objects, but this step is unneeded too.
>>>
>>> Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
>>> Tested-by: Nathan Chancellor <nathan@kernel.org>
>>> Tested-by: Nicolas Schier <nicolas@fjasle.eu>
>>> Reviewed-by: Nicolas Schier <nicolas@fjasle.eu>
>>
>> Problem with v6.0-rc1
>> Problem with v5.19
>> No problem with v5.18
>>
>> Bisected to 7b4537199a4a ("kbuild: link symbol CRCs at final link,
>> removing CONFIG_MODULE_REL_CRCS")
>>
>> The above patch leads to the following problem building
>> mpc85xx_defconfig + CONFIG_RELOCATABLE
> 
> 
> 
> Is this because the relocation implementation on ppc is incomplete?
> (and is it the reason why relock_check.sh exists?)
> 
> arch/powerpc/kernel/reloc_32.S does not support R_PPC_UADDR32
> 
> 

Might be the reason.

Is it expected that your patch adds an unsupported relocation ?

Why was that relocation type unneeded before ?

Thanks
Christophe
Masahiro Yamada Aug. 20, 2022, 5:01 p.m. UTC | #6
On Sat, Aug 20, 2022 at 11:15 PM Christophe Leroy
<christophe.leroy@csgroup.eu> wrote:
>
>
>
> Le 20/08/2022 à 14:51, Masahiro Yamada a écrit :
> > On Sat, Aug 20, 2022 at 7:02 PM Christophe Leroy
> > <christophe.leroy@csgroup.eu> wrote:
> >>
> >> Hi,
> >>
> >> Le 13/05/2022 à 13:39, Masahiro Yamada a écrit :
> >>> include/{linux,asm-generic}/export.h defines a weak symbol, __crc_*
> >>> as a placeholder.
> >>>
> >>> Genksyms writes the version CRCs into the linker script, which will be
> >>> used for filling the __crc_* symbols. The linker script format depends
> >>> on CONFIG_MODULE_REL_CRCS. If it is enabled, __crc_* holds the offset
> >>> to the reference of CRC.
> >>>
> >>> It is time to get rid of this complexity.
> >>>
> >>> Now that modpost parses text files (.*.cmd) to collect all the CRCs,
> >>> it can generate C code that will be linked to the vmlinux or modules.
> >>>
> >>> Generate a new C file, .vmlinux.export.c, which contains the CRCs of
> >>> symbols exported by vmlinux. It is compiled and linked to vmlinux in
> >>> scripts/link-vmlinux.sh.
> >>>
> >>> Put the CRCs of symbols exported by modules into the existing *.mod.c
> >>> files. No additional build step is needed for modules. As before,
> >>> *.mod.c are compiled and linked to *.ko in scripts/Makefile.modfinal.
> >>>
> >>> No linker magic is used here. The new C implementation works in the
> >>> same way, whether CONFIG_RELOCATABLE is enabled or not.
> >>> CONFIG_MODULE_REL_CRCS is no longer needed.
> >>>
> >>> Previously, Kbuild invoked additional $(LD) to update the CRCs in
> >>> objects, but this step is unneeded too.
> >>>
> >>> Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
> >>> Tested-by: Nathan Chancellor <nathan@kernel.org>
> >>> Tested-by: Nicolas Schier <nicolas@fjasle.eu>
> >>> Reviewed-by: Nicolas Schier <nicolas@fjasle.eu>
> >>
> >> Problem with v6.0-rc1
> >> Problem with v5.19
> >> No problem with v5.18
> >>
> >> Bisected to 7b4537199a4a ("kbuild: link symbol CRCs at final link,
> >> removing CONFIG_MODULE_REL_CRCS")
> >>
> >> The above patch leads to the following problem building
> >> mpc85xx_defconfig + CONFIG_RELOCATABLE
> >
> >
> >
> > Is this because the relocation implementation on ppc is incomplete?
> > (and is it the reason why relock_check.sh exists?)
> >
> > arch/powerpc/kernel/reloc_32.S does not support R_PPC_UADDR32
> >
> >
>
> Might be the reason.
>
> Is it expected that your patch adds an unsupported relocation ?
>
> Why was that relocation type unneeded before ?
>
> Thanks
> Christophe


I posted a patch (although I believe my commit is innocent).

https://lore.kernel.org/lkml/20220820165129.1147589-1-masahiroy@kernel.org/T/#u

The relocs_check.sh warnings are gone.
Please do a boot test.
Thanks.
Christophe Leroy Aug. 20, 2022, 5:51 p.m. UTC | #7
Le 20/08/2022 à 19:01, Masahiro Yamada a écrit :
> On Sat, Aug 20, 2022 at 11:15 PM Christophe Leroy
> <christophe.leroy@csgroup.eu> wrote:
>>
>>
>>
>> Le 20/08/2022 à 14:51, Masahiro Yamada a écrit :
>>> On Sat, Aug 20, 2022 at 7:02 PM Christophe Leroy
>>> <christophe.leroy@csgroup.eu> wrote:
>>>>
>>>> Hi,
>>>>
>>>> Le 13/05/2022 à 13:39, Masahiro Yamada a écrit :
>>>>> include/{linux,asm-generic}/export.h defines a weak symbol, __crc_*
>>>>> as a placeholder.
>>>>>
>>>>> Genksyms writes the version CRCs into the linker script, which will be
>>>>> used for filling the __crc_* symbols. The linker script format depends
>>>>> on CONFIG_MODULE_REL_CRCS. If it is enabled, __crc_* holds the offset
>>>>> to the reference of CRC.
>>>>>
>>>>> It is time to get rid of this complexity.
>>>>>
>>>>> Now that modpost parses text files (.*.cmd) to collect all the CRCs,
>>>>> it can generate C code that will be linked to the vmlinux or modules.
>>>>>
>>>>> Generate a new C file, .vmlinux.export.c, which contains the CRCs of
>>>>> symbols exported by vmlinux. It is compiled and linked to vmlinux in
>>>>> scripts/link-vmlinux.sh.
>>>>>
>>>>> Put the CRCs of symbols exported by modules into the existing *.mod.c
>>>>> files. No additional build step is needed for modules. As before,
>>>>> *.mod.c are compiled and linked to *.ko in scripts/Makefile.modfinal.
>>>>>
>>>>> No linker magic is used here. The new C implementation works in the
>>>>> same way, whether CONFIG_RELOCATABLE is enabled or not.
>>>>> CONFIG_MODULE_REL_CRCS is no longer needed.
>>>>>
>>>>> Previously, Kbuild invoked additional $(LD) to update the CRCs in
>>>>> objects, but this step is unneeded too.
>>>>>
>>>>> Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
>>>>> Tested-by: Nathan Chancellor <nathan@kernel.org>
>>>>> Tested-by: Nicolas Schier <nicolas@fjasle.eu>
>>>>> Reviewed-by: Nicolas Schier <nicolas@fjasle.eu>
>>>>
>>>> Problem with v6.0-rc1
>>>> Problem with v5.19
>>>> No problem with v5.18
>>>>
>>>> Bisected to 7b4537199a4a ("kbuild: link symbol CRCs at final link,
>>>> removing CONFIG_MODULE_REL_CRCS")
>>>>
>>>> The above patch leads to the following problem building
>>>> mpc85xx_defconfig + CONFIG_RELOCATABLE
>>>
>>>
>>>
>>> Is this because the relocation implementation on ppc is incomplete?
>>> (and is it the reason why relock_check.sh exists?)
>>>
>>> arch/powerpc/kernel/reloc_32.S does not support R_PPC_UADDR32
>>>
>>>
>>
>> Might be the reason.
>>
>> Is it expected that your patch adds an unsupported relocation ?
>>
>> Why was that relocation type unneeded before ?
>>
>> Thanks
>> Christophe
> 
> 
> I posted a patch (although I believe my commit is innocent).
> 
> https://lore.kernel.org/lkml/20220820165129.1147589-1-masahiroy@kernel.org/T/#u
> 
> The relocs_check.sh warnings are gone.
> Please do a boot test.
> Thanks.
> 

Yes it works, many Thanks.

The fixes tag should probably be c857c43b34ec ("powerpc: Don't use a 
function descriptor for system call table")


Christophe
diff mbox series

Patch

diff --git a/arch/m68k/include/asm/Kbuild b/arch/m68k/include/asm/Kbuild
index 0dbf9c5c6fae..1b720299deb1 100644
--- a/arch/m68k/include/asm/Kbuild
+++ b/arch/m68k/include/asm/Kbuild
@@ -1,5 +1,6 @@ 
 # SPDX-License-Identifier: GPL-2.0
 generated-y += syscall_table.h
+generic-y += export.h
 generic-y += extable.h
 generic-y += kvm_para.h
 generic-y += mcs_spinlock.h
diff --git a/arch/m68k/include/asm/export.h b/arch/m68k/include/asm/export.h
deleted file mode 100644
index b53008b67ce1..000000000000
--- a/arch/m68k/include/asm/export.h
+++ /dev/null
@@ -1,2 +0,0 @@ 
-#define KCRC_ALIGN 2
-#include <asm-generic/export.h>
diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig
index 174edabb74fa..a4e8dd889e29 100644
--- a/arch/powerpc/Kconfig
+++ b/arch/powerpc/Kconfig
@@ -566,7 +566,6 @@  config RELOCATABLE
 	bool "Build a relocatable kernel"
 	depends on PPC64 || (FLATMEM && (44x || FSL_BOOKE))
 	select NONSTATIC_KERNEL
-	select MODULE_REL_CRCS if MODVERSIONS
 	help
 	  This builds a kernel image that is capable of running at the
 	  location the kernel is loaded at. For ppc32, there is no any
diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig
index 77b5a03de13a..aa5848004c76 100644
--- a/arch/s390/Kconfig
+++ b/arch/s390/Kconfig
@@ -567,7 +567,6 @@  endchoice
 
 config RELOCATABLE
 	bool "Build a relocatable kernel"
-	select MODULE_REL_CRCS if MODVERSIONS
 	default y
 	help
 	  This builds a kernel image that retains relocation information
diff --git a/arch/um/Kconfig b/arch/um/Kconfig
index 4d398b80aea8..e8983d098e73 100644
--- a/arch/um/Kconfig
+++ b/arch/um/Kconfig
@@ -106,7 +106,6 @@  config LD_SCRIPT_DYN
 	bool
 	default y
 	depends on !LD_SCRIPT_STATIC
-	select MODULE_REL_CRCS if MODVERSIONS
 
 config LD_SCRIPT_DYN_RPATH
 	bool "set rpath in the binary" if EXPERT
diff --git a/include/asm-generic/export.h b/include/asm-generic/export.h
index 07a36a874dca..5e4b1f2369d2 100644
--- a/include/asm-generic/export.h
+++ b/include/asm-generic/export.h
@@ -2,6 +2,14 @@ 
 #ifndef __ASM_GENERIC_EXPORT_H
 #define __ASM_GENERIC_EXPORT_H
 
+/*
+ * This comment block is used by fixdep. Please do not remove.
+ *
+ * When CONFIG_MODVERSIONS is changed from n to y, all source files having
+ * EXPORT_SYMBOL variants must be re-compiled because genksyms is run as a
+ * side effect of the *.o build rule.
+ */
+
 #ifndef KSYM_FUNC
 #define KSYM_FUNC(x) x
 #endif
@@ -12,9 +20,6 @@ 
 #else
 #define KSYM_ALIGN 4
 #endif
-#ifndef KCRC_ALIGN
-#define KCRC_ALIGN 4
-#endif
 
 .macro __put, val, name
 #ifdef CONFIG_HAVE_ARCH_PREL32_RELOCATIONS
@@ -43,17 +48,6 @@  __ksymtab_\name:
 __kstrtab_\name:
 	.asciz "\name"
 	.previous
-#ifdef CONFIG_MODVERSIONS
-	.section ___kcrctab\sec+\name,"a"
-	.balign KCRC_ALIGN
-#if defined(CONFIG_MODULE_REL_CRCS)
-	.long __crc_\name - .
-#else
-	.long __crc_\name
-#endif
-	.weak __crc_\name
-	.previous
-#endif
 #endif
 .endm
 
diff --git a/include/linux/export-internal.h b/include/linux/export-internal.h
new file mode 100644
index 000000000000..77175d561058
--- /dev/null
+++ b/include/linux/export-internal.h
@@ -0,0 +1,16 @@ 
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Please do not include this explicitly.
+ * This is used by C files generated by modpost.
+ */
+
+#ifndef __LINUX_EXPORT_INTERNAL_H__
+#define __LINUX_EXPORT_INTERNAL_H__
+
+#include <linux/compiler.h>
+#include <linux/types.h>
+
+#define SYMBOL_CRC(sym, crc, sec)   \
+	u32 __section("___kcrctab" sec "+" #sym) __crc_##sym = crc
+
+#endif /* __LINUX_EXPORT_INTERNAL_H__ */
diff --git a/include/linux/export.h b/include/linux/export.h
index 27d848712b90..565c5ffcb26f 100644
--- a/include/linux/export.h
+++ b/include/linux/export.h
@@ -11,6 +11,14 @@ 
  * hackers place grumpy comments in header files.
  */
 
+/*
+ * This comment block is used by fixdep. Please do not remove.
+ *
+ * When CONFIG_MODVERSIONS is changed from n to y, all source files having
+ * EXPORT_SYMBOL variants must be re-compiled because genksyms is run as a
+ * side effect of the *.o build rule.
+ */
+
 #ifndef __ASSEMBLY__
 #ifdef MODULE
 extern struct module __this_module;
@@ -19,26 +27,6 @@  extern struct module __this_module;
 #define THIS_MODULE ((struct module *)0)
 #endif
 
-#ifdef CONFIG_MODVERSIONS
-/* Mark the CRC weak since genksyms apparently decides not to
- * generate a checksums for some symbols */
-#if defined(CONFIG_MODULE_REL_CRCS)
-#define __CRC_SYMBOL(sym, sec)						\
-	asm("	.section \"___kcrctab" sec "+" #sym "\", \"a\"	\n"	\
-	    "	.weak	__crc_" #sym "				\n"	\
-	    "	.long	__crc_" #sym " - .			\n"	\
-	    "	.previous					\n")
-#else
-#define __CRC_SYMBOL(sym, sec)						\
-	asm("	.section \"___kcrctab" sec "+" #sym "\", \"a\"	\n"	\
-	    "	.weak	__crc_" #sym "				\n"	\
-	    "	.long	__crc_" #sym "				\n"	\
-	    "	.previous					\n")
-#endif
-#else
-#define __CRC_SYMBOL(sym, sec)
-#endif
-
 #ifdef CONFIG_HAVE_ARCH_PREL32_RELOCATIONS
 #include <linux/compiler.h>
 /*
@@ -85,7 +73,6 @@  struct kernel_symbol {
 /*
  * For every exported symbol, do the following:
  *
- * - If applicable, place a CRC entry in the __kcrctab section.
  * - Put the name of the symbol and namespace (empty string "" for none) in
  *   __ksymtab_strings.
  * - Place a struct kernel_symbol entry in the __ksymtab section.
@@ -98,7 +85,6 @@  struct kernel_symbol {
 	extern typeof(sym) sym;							\
 	extern const char __kstrtab_##sym[];					\
 	extern const char __kstrtabns_##sym[];					\
-	__CRC_SYMBOL(sym, sec);							\
 	asm("	.section \"__ksymtab_strings\",\"aMS\",%progbits,1	\n"	\
 	    "__kstrtab_" #sym ":					\n"	\
 	    "	.asciz 	\"" #sym "\"					\n"	\
diff --git a/init/Kconfig b/init/Kconfig
index ddcbefe535e9..f5b14318dfcb 100644
--- a/init/Kconfig
+++ b/init/Kconfig
@@ -2136,10 +2136,6 @@  config ASM_MODVERSIONS
 	  assembly. This can be enabled only when the target architecture
 	  supports it.
 
-config MODULE_REL_CRCS
-	bool
-	depends on MODVERSIONS
-
 config MODULE_SRCVERSION_ALL
 	bool "Source checksum for all modules"
 	help
diff --git a/kernel/module.c b/kernel/module.c
index 6cea788fd965..c9e2342da28e 100644
--- a/kernel/module.c
+++ b/kernel/module.c
@@ -1231,11 +1231,6 @@  static int try_to_force_load(struct module *mod, const char *reason)
 
 #ifdef CONFIG_MODVERSIONS
 
-static u32 resolve_rel_crc(const s32 *crc)
-{
-	return *(u32 *)((void *)crc + *crc);
-}
-
 static int check_version(const struct load_info *info,
 			 const char *symname,
 			 struct module *mod,
@@ -1264,10 +1259,7 @@  static int check_version(const struct load_info *info,
 		if (strcmp(versions[i].name, symname) != 0)
 			continue;
 
-		if (IS_ENABLED(CONFIG_MODULE_REL_CRCS))
-			crcval = resolve_rel_crc(crc);
-		else
-			crcval = *crc;
+		crcval = *crc;
 		if (versions[i].crc == crcval)
 			return 1;
 		pr_debug("Found checksum %X vs module %lX\n",
diff --git a/scripts/Makefile.build b/scripts/Makefile.build
index a1023868775f..ddd9080fc028 100644
--- a/scripts/Makefile.build
+++ b/scripts/Makefile.build
@@ -128,7 +128,6 @@  $(obj)/%.i: $(src)/%.c FORCE
 
 genksyms = scripts/genksyms/genksyms		\
 	$(if $(1), -T $(2))			\
-	$(if $(CONFIG_MODULE_REL_CRCS), -R)	\
 	$(if $(KBUILD_PRESERVE), -p)		\
 	-r $(or $(wildcard $(2:.symtypes=.symref)), /dev/null)
 
@@ -162,19 +161,11 @@  ifdef CONFIG_MODVERSIONS
 # o if <file>.o doesn't contain a __ksymtab version, i.e. does
 #   not export symbols, it's done.
 # o otherwise, we calculate symbol versions using the good old
-#   genksyms on the preprocessed source and postprocess them in a way
-#   that they are usable as a linker script
-# o generate .tmp_<file>.o from <file>.o using the linker to
-#   replace the unresolved symbols __crc_exported_symbol with
-#   the actual value of the checksum generated by genksyms
-# o remove .tmp_<file>.o to <file>.o
+#   genksyms on the preprocessed source and dump them into the .cmd file.
+# o modpost will extract versions from that file and create *.c files that will
+#   be compiled and linked to the kernel and/or modules.
 
-# Generate .o.symversions files for each .o with exported symbols, and link these
-# to the kernel and/or modules at the end.
-
-genksyms_format_rel_crc := [^_]*__crc_\([^ ]*\) = \.; LONG(\([^)]*\)).*
-genksyms_format_normal := __crc_\(.*\) = \(.*\);
-genksyms_format := $(if $(CONFIG_MODULE_REL_CRCS),$(genksyms_format_rel_crc),$(genksyms_format_normal))
+genksyms_format := __crc_\(.*\) = \(.*\);
 
 gen_symversions =								\
 	if $(NM) $@ 2>/dev/null | grep -q __ksymtab; then			\
@@ -188,12 +179,6 @@  gen_symversions =								\
 
 cmd_gen_symversions_c =	$(call gen_symversions,c)
 
-cmd_modversions =								\
-	if [ -r $@.symversions ]; then						\
-		$(LD) $(KBUILD_LDFLAGS) -r -o $(@D)/.tmp_$(@F) $@ 		\
-			-T $@.symversions;					\
-		mv -f $(@D)/.tmp_$(@F) $@;					\
-	fi
 endif
 
 ifdef CONFIG_FTRACE_MCOUNT_USE_RECORDMCOUNT
@@ -273,7 +258,6 @@  define rule_cc_o_c
 	$(call cmd,checkdoc)
 	$(call cmd,gen_objtooldep)
 	$(call cmd,gen_symversions_c)
-	$(if $(CONFIG_LTO_CLANG),,$(call cmd,modversions))
 	$(call cmd,record_mcount)
 endef
 
@@ -282,7 +266,6 @@  define rule_as_o_S
 	$(call cmd,gen_ksymdeps)
 	$(call cmd,gen_objtooldep)
 	$(call cmd,gen_symversions_S)
-	$(call cmd,modversions)
 endef
 
 # Built-in and composite module parts
@@ -296,8 +279,6 @@  ifneq ($(CONFIG_LTO_CLANG)$(CONFIG_X86_KERNEL_IBT),)
 quiet_cmd_cc_prelink_modules = LD [M]  $@
       cmd_cc_prelink_modules =						\
 	$(LD) $(ld_flags) -r -o $@					\
-               $(shell [ -s $(@:.prelink.o=.o.symversions) ] &&		\
-                       echo -T $(@:.prelink.o=.o.symversions))		\
 		--whole-archive $(filter-out FORCE,$^)			\
 		$(cmd_objtool)
 
diff --git a/scripts/genksyms/genksyms.c b/scripts/genksyms/genksyms.c
index 4827c5abe5b7..67b23cc0df0f 100644
--- a/scripts/genksyms/genksyms.c
+++ b/scripts/genksyms/genksyms.c
@@ -33,7 +33,7 @@  char *cur_filename;
 int in_source_file;
 
 static int flag_debug, flag_dump_defs, flag_reference, flag_dump_types,
-	   flag_preserve, flag_warnings, flag_rel_crcs;
+	   flag_preserve, flag_warnings;
 
 static int errors;
 static int nsyms;
@@ -680,11 +680,7 @@  void export_symbol(const char *name)
 		if (flag_dump_defs)
 			fputs(">\n", debugfile);
 
-		/* Used as a linker script. */
-		printf(!flag_rel_crcs ? "__crc_%s = 0x%08lx;\n" :
-		       "SECTIONS { .rodata : ALIGN(4) { "
-		       "__crc_%s = .; LONG(0x%08lx); } }\n",
-		       name, crc);
+		printf("__crc_%s = 0x%08lx;\n", name, crc);
 	}
 }
 
@@ -733,7 +729,6 @@  static void genksyms_usage(void)
 	      "  -q, --quiet           Disable warnings (default)\n"
 	      "  -h, --help            Print this message\n"
 	      "  -V, --version         Print the release version\n"
-	      "  -R, --relative-crc    Emit section relative symbol CRCs\n"
 #else				/* __GNU_LIBRARY__ */
 	      "  -s                    Select symbol prefix\n"
 	      "  -d                    Increment the debug level (repeatable)\n"
@@ -745,7 +740,6 @@  static void genksyms_usage(void)
 	      "  -q                    Disable warnings (default)\n"
 	      "  -h                    Print this message\n"
 	      "  -V                    Print the release version\n"
-	      "  -R                    Emit section relative symbol CRCs\n"
 #endif				/* __GNU_LIBRARY__ */
 	      , stderr);
 }
@@ -766,14 +760,13 @@  int main(int argc, char **argv)
 		{"preserve", 0, 0, 'p'},
 		{"version", 0, 0, 'V'},
 		{"help", 0, 0, 'h'},
-		{"relative-crc", 0, 0, 'R'},
 		{0, 0, 0, 0}
 	};
 
-	while ((o = getopt_long(argc, argv, "s:dwqVDr:T:phR",
+	while ((o = getopt_long(argc, argv, "s:dwqVDr:T:ph",
 				&long_opts[0], NULL)) != EOF)
 #else				/* __GNU_LIBRARY__ */
-	while ((o = getopt(argc, argv, "s:dwqVDr:T:phR")) != EOF)
+	while ((o = getopt(argc, argv, "s:dwqVDr:T:ph")) != EOF)
 #endif				/* __GNU_LIBRARY__ */
 		switch (o) {
 		case 'd':
@@ -813,9 +806,6 @@  int main(int argc, char **argv)
 		case 'h':
 			genksyms_usage();
 			return 0;
-		case 'R':
-			flag_rel_crcs = 1;
-			break;
 		default:
 			genksyms_usage();
 			return 1;
diff --git a/scripts/link-vmlinux.sh b/scripts/link-vmlinux.sh
index eceb3ee7ec06..7313cbd755df 100755
--- a/scripts/link-vmlinux.sh
+++ b/scripts/link-vmlinux.sh
@@ -90,7 +90,6 @@  modpost_link()
 
 		if is_enabled CONFIG_MODVERSIONS; then
 			gen_symversions
-			lds="${lds} -T .tmp_symversions.lds"
 		fi
 
 		# This might take a while, so indicate that we're doing
@@ -183,6 +182,10 @@  vmlinux_link()
 		libs="${KBUILD_VMLINUX_LIBS}"
 	fi
 
+	if is_enabled CONFIG_MODULES; then
+		objs="${objs} .vmlinux.export.o"
+	fi
+
 	if [ "${SRCARCH}" = "um" ]; then
 		wl=-Wl,
 		ld="${CC}"
@@ -312,6 +315,7 @@  cleanup()
 	rm -f vmlinux.o
 	rm -f .vmlinux.d
 	rm -f .vmlinux.objs
+	rm -f .vmlinux.export.c
 }
 
 # Use "make V=1" to debug this script
@@ -363,6 +367,14 @@  info GEN modules.builtin
 tr '\0' '\n' < modules.builtin.modinfo | sed -n 's/^[[:alnum:]:_]*\.file=//p' |
 	tr ' ' '\n' | uniq | sed -e 's:^:kernel/:' -e 's/$/.ko/' > modules.builtin
 
+if is_enabled CONFIG_MODULES; then
+	info CC .vmlinux.export.c
+	${CC} ${NOSTDINC_FLAGS} ${LINUXINCLUDE} \
+	      ${KBUILD_CPPFLAGS} ${KBUILD_CFLAGS} \
+	      ${KBUILD_CFLAGS_KERNEL} ${CFLAGS_KERNEL} \
+	      -c -o .vmlinux.export.o .vmlinux.export.c
+fi
+
 btf_vmlinux_bin_o=""
 if is_enabled CONFIG_DEBUG_INFO_BTF; then
 	btf_vmlinux_bin_o=.btf.vmlinux.bin.o
diff --git a/scripts/mod/modpost.c b/scripts/mod/modpost.c
index 213a130d1635..2313bdae91bf 100644
--- a/scripts/mod/modpost.c
+++ b/scripts/mod/modpost.c
@@ -1042,6 +1042,7 @@  static void add_header(struct buffer *b, struct module *mod)
 	buf_printf(b, "#define INCLUDE_VERMAGIC\n");
 	buf_printf(b, "#include <linux/build-salt.h>\n");
 	buf_printf(b, "#include <linux/elfnote-lto.h>\n");
+	buf_printf(b, "#include <linux/export-internal.h>\n");
 	buf_printf(b, "#include <linux/vermagic.h>\n");
 	buf_printf(b, "#include <linux/compiler.h>\n");
 	buf_printf(b, "\n");
@@ -1076,20 +1077,26 @@  static void add_header(struct buffer *b, struct module *mod)
 		buf_printf(b, "\nMODULE_INFO(staging, \"Y\");\n");
 }
 
-static void check_symversions(struct module *mod)
+static void add_exported_symbols(struct buffer *buf, struct module *mod)
 {
 	struct symbol *sym;
 
 	if (!modversions)
 		return;
 
+	/* record CRCs for exported symbols */
+	buf_printf(buf, "\n");
 	list_for_each_entry(sym, &mod->exported_symbols, list) {
 		if (!sym->crc_valid) {
 			warn("EXPORT symbol \"%s\" [%s%s] version generation failed, symbol will not be versioned.\n"
 			     "Is \"%s\" prototyped in <asm/asm-prototypes.h>?\n",
 			     sym->name, mod->name, mod->is_vmlinux ? "" : ".ko",
 			     sym->name);
+			continue;
 		}
+
+		buf_printf(buf, "SYMBOL_CRC(%s, 0x%08x, \"%s\");\n",
+			   sym->name, sym->crc, sym->is_gpl_only ? "_gpl" : "");
 	}
 }
 
@@ -1226,6 +1233,18 @@  static void write_if_changed(struct buffer *b, const char *fname)
 	write_buf(b, fname);
 }
 
+static void write_vmlinux_export_c_file(struct module *mod)
+{
+	struct buffer buf = { };
+
+	buf_printf(&buf,
+		   "#include <linux/export-internal.h>\n");
+
+	add_exported_symbols(&buf, mod);
+	write_if_changed(&buf, ".vmlinux.export.c");
+	free(buf.p);
+}
+
 /* do sanity checks, and generate *.mod.c file */
 static void write_mod_c_file(struct module *mod)
 {
@@ -1237,6 +1256,7 @@  static void write_mod_c_file(struct module *mod)
 	check_exports(mod);
 
 	add_header(&buf, mod);
+	add_exported_symbols(&buf, mod);
 	add_versions(&buf, mod);
 	add_depends(&buf, mod);
 	add_moddevtable(&buf, mod);
@@ -1434,9 +1454,9 @@  int main(int argc, char **argv)
 		if (mod->from_dump)
 			continue;
 
-		check_symversions(mod);
-
-		if (!mod->is_vmlinux)
+		if (mod->is_vmlinux)
+			write_vmlinux_export_c_file(mod);
+		else
 			write_mod_c_file(mod);
 	}