When playing with gendwarfksyms v6 from Sami Tolvanen I noticed: $ LC_ALL=C ls -alth .tmp_vmlinux* -rw-rw-r-- 1 dileks dileks 3.1M Nov 22 20:52 .tmp_vmlinux2.kallsyms.o -rw-rw-r-- 1 dileks dileks 34M Nov 22 20:52 .tmp_vmlinux2.kallsyms.S -rw-rw-r-- 1 dileks dileks 6.5M Nov 22 20:52 .tmp_vmlinux2.syms -rwxrwxr-x 1 dileks dileks 101M Nov 22 20:52 .tmp_vmlinux2 -rw-rw-r-- 1 dileks dileks 3.1M Nov 22 20:52 .tmp_vmlinux1.kallsyms.o -rw-rw-r-- 1 dileks dileks 34M Nov 22 20:52 .tmp_vmlinux1.kallsyms.S -rw-rw-r-- 1 dileks dileks 6.5M Nov 22 20:52 .tmp_vmlinux1.syms -rwxrwxr-x 1 dileks dileks 52M Nov 22 20:52 .tmp_vmlinux1.btf.o -rwxrwxr-x 1 dileks dileks 514M Nov 22 20:52 .tmp_vmlinux1 -rw-rw-r-- 1 dileks dileks 2.1K Nov 22 20:51 .tmp_vmlinux0.kallsyms.o -rw-rw-r-- 1 dileks dileks 6.3K Nov 22 20:51 .tmp_vmlinux0.kallsyms.S -rw-rw-r-- 1 dileks dileks 0 Nov 22 20:51 .tmp_vmlinux.kallsyms0.syms .tmp_vmlinux.kallsyms0.syms is NULL byte - it's a dummy file. Further looking at the other .tmp_vmlinux syms files: .tmp_vmlinux2.syms .tmp_vmlinux1.syms .tmp_vmlinux.kallsyms0.syms Change the naming of file .tmp_vmlinux.kallsyms0.syms -> .tmp_vmlinux0.syms While at this, fix the comments in scripts/link-vmlinux.sh. INFO: v2 is based on Linux v6.12 Link: https://github.com/samitolvanen/linux/commits/gendwarfksyms-v6 Link: https://lore.kernel.org/all/CA+icZUXvu0Kw8RH1ZGBKgYGG-8u9x8BbsEkjtm4vSVKkXPTg+Q@xxxxxxxxxxxxxx/ Link: https://lore.kernel.org/all/20241123132237.15700-1-sedat.dilek@xxxxxxxxx/ (v1) Signed-off-by: Sedat Dilek <sedat.dilek@xxxxxxxxx> --- v1 -> v2: - Add commit description and follow naming consistency as requested by Masahiroy san. -dileks scripts/link-vmlinux.sh | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/scripts/link-vmlinux.sh b/scripts/link-vmlinux.sh index a9b3f34a78d2..239fe036606f 100755 --- a/scripts/link-vmlinux.sh +++ b/scripts/link-vmlinux.sh @@ -203,8 +203,8 @@ kallsymso= strip_debug= if is_enabled CONFIG_KALLSYMS; then - true > .tmp_vmlinux.kallsyms0.syms - kallsyms .tmp_vmlinux.kallsyms0.syms .tmp_vmlinux0.kallsyms + true > .tmp_vmlinux0.syms + kallsyms .tmp_vmlinux0.syms .tmp_vmlinux0.kallsyms fi if is_enabled CONFIG_KALLSYMS || is_enabled CONFIG_DEBUG_INFO_BTF; then @@ -231,14 +231,14 @@ if is_enabled CONFIG_KALLSYMS; then # Generate section listing all symbols and add it into vmlinux # It's a four step process: # 0) Generate a dummy __kallsyms with empty symbol list. - # 1) Link .tmp_vmlinux.kallsyms1 so it has all symbols and sections, + # 1) Link .tmp_vmlinux1.kallsyms so it has all symbols and sections, # with a dummy __kallsyms. - # Running kallsyms on that gives us .tmp_kallsyms1.o with + # Running kallsyms on that gives us .tmp_vmlinux1.kallsyms.o with # the right size - # 2) Link .tmp_vmlinux.kallsyms2 so it now has a __kallsyms section of + # 2) Link .tmp_vmlinux2.kallsyms so it now has a __kallsyms section of # the right size, but due to the added section, some # addresses have shifted. - # From here, we generate a correct .tmp_vmlinux.kallsyms2.o + # From here, we generate a correct .tmp_vmlinux2.kallsyms.o # 3) That link may have expanded the kernel image enough that # more linker branch stubs / trampolines had to be added, which # introduces new names, which further expands kallsyms. Do another -- 2.45.2
