Revision 4a33d4f17a19ece50ccf5ccbb8bc4ee539fb492e authored by Masahiro Yamada on 20 May 2019, 02:54:37 UTC, committed by Masahiro Yamada on 20 May 2019, 23:54:05 UTC
I just thought it was a good idea to scan builtin.modules in the name uniqueness checking, but a couple of false positives were found. Stephen reported a false positive for ppc64_defconfig: warning: same basename if the following are built as modules: arch/powerpc/platforms/powermac/nvram.ko drivers/char/nvram.ko The former is never built as a module as you see in arch/powerpc/platforms/powermac/Makefile: # CONFIG_NVRAM is an arch. independent tristate symbol, for pmac32 we really # need this to be a bool. Cheat here and pretend CONFIG_NVRAM=m is really # CONFIG_NVRAM=y obj-$(CONFIG_NVRAM:m=y) += nvram.o Another example of false positive is arm64 defconfig: warning: same basename if the following are built as modules: arch/arm64/lib/crc32.ko lib/crc32.ko It is true CONFIG_CRC32 is a tristate option but it is always 'y' since it is select'ed by ARM64. Hence, neither of them is built as a module for the arm64 build. From the above, modules.builtin essentially contains false positives. I do not think it is a big deal as far as kmod is concerned, but false positive warnings in the kernel build make people upset. It is better to not check it. Even without builtin.modules checked, we have enough (and more solid) test coverage with allmodconfig. While I touched this part, I replaced the sed code with neater one provided by Stephen. Link: https://lkml.org/lkml/2019/5/19/120 Link: https://lkml.org/lkml/2019/5/19/123 Fixes: 3a48a91901c5 ("kbuild: check uniqueness of module names") Reported-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> Acked-by: Arnd Bergmann <arnd@arndb.de> Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
1 parent 055efab
mksysmap
#!/bin/sh -x
# Based on the vmlinux file create the System.map file
# System.map is used by module-init tools and some debugging
# tools to retrieve the actual addresses of symbols in the kernel.
#
# Usage
# mksysmap vmlinux System.map
#####
# Generate System.map (actual filename passed as second argument)
# $NM produces the following output:
# f0081e80 T alloc_vfsmnt
# The second row specify the type of the symbol:
# A = Absolute
# B = Uninitialised data (.bss)
# C = Common symbol
# D = Initialised data
# G = Initialised data for small objects
# I = Indirect reference to another symbol
# N = Debugging symbol
# R = Read only
# S = Uninitialised data for small objects
# T = Text code symbol
# U = Undefined symbol
# V = Weak symbol
# W = Weak symbol
# Corresponding small letters are local symbols
# For System.map filter away:
# a - local absolute symbols
# U - undefined global symbols
# N - debugging symbols
# w - local weak symbols
# readprofile starts reading symbols when _stext is found, and
# continue until it finds a symbol which is not either of 'T', 't',
# 'W' or 'w'. __crc_ are 'A' and placed in the middle
# so we just ignore them to let readprofile continue to work.
# (At least sparc64 has __crc_ in the middle).
$NM -n $1 | grep -v '\( [aNUw] \)\|\(__crc_\)\|\( \$[adt]\)\|\( .L\)' > $2
![swh spinner](/static/img/swh-spinner.gif)
Computing file changes ...