scripts/kallsyms: skip ppc compiler stub *.long_branch.* / *.plt_branch.*
authorMasahiro Yamada <masahiroy@kernel.org>
Tue, 22 Sep 2020 17:48:56 +0000 (02:48 +0900)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 7 Oct 2020 06:02:50 +0000 (08:02 +0200)
commit6bef7d4b4770c42bb4abd5215801121672f7acd9
tree887fa6172fa76c89d14f16e07371607d779e148b
parenta50ea89d1ae546bcd28a7afb0be45cdb1a1138ba
scripts/kallsyms: skip ppc compiler stub *.long_branch.* / *.plt_branch.*

[ Upstream commit 516d980f85415d76ae3d0d2a871eb20243f46c95 ]

PowerPC allmodconfig often fails to build as follows:

    LD      .tmp_vmlinux.kallsyms1
    KSYM    .tmp_vmlinux.kallsyms1.o
    LD      .tmp_vmlinux.kallsyms2
    KSYM    .tmp_vmlinux.kallsyms2.o
    LD      .tmp_vmlinux.kallsyms3
    KSYM    .tmp_vmlinux.kallsyms3.o
    LD      vmlinux
    SORTTAB vmlinux
    SYSMAP  System.map
  Inconsistent kallsyms data
  Try make KALLSYMS_EXTRA_PASS=1 as a workaround
  make[2]: *** [../Makefile:1162: vmlinux] Error 1

Setting KALLSYMS_EXTRA_PASS=1 does not help.

This is caused by the compiler inserting stubs such as *.long_branch.*
and *.plt_branch.*

  $ powerpc-linux-nm -n .tmp_vmlinux.kallsyms2
   [ snip ]
  c00000000210c010 t 00000075.plt_branch.da9:19
  c00000000210c020 t 00000075.plt_branch.1677:5
  c00000000210c030 t 00000075.long_branch.memmove
  c00000000210c034 t 00000075.plt_branch.9e0:5
  c00000000210c044 t 00000075.plt_branch.free_initrd_mem
    ...

Actually, the problem mentioned in scripts/link-vmlinux.sh comments;
"In theory it's possible this results in even more stubs, but unlikely"
is happening here, and ends up with another kallsyms step required.

scripts/kallsyms.c already ignores various compiler stubs. Let's do
similar to make kallsysms for PowerPC always succeed in 2 steps.

Reported-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Tested-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
scripts/kallsyms.c