4fcfdefddb
It leads to some regressions in requires generation precision, though,
as it is no longer possible to exclude requires that are satisfied
by other kmods in the package. And calling a script on each file (instead
of calling it on the whole file list at once) does not make things faster,
either (so much for "sanity and benefit").
* find-provides.ksyms: Check for "$@" as well.
* find-requires.ksyms: Likewise.
* kernel-srpm-macros.spec (Source104, Source105, Source106): New
attribute files.
(%install): Install provided_ksyms.attr, required_ksyms.attr,
and modalias.attr into "%{buildroot}%{_fileattrsdir}".
(%files -n kernel-rpm-macros): Add provided_ksyms.attr,
required_ksyms.attr, and modalias.attr.
kmodtool: Remove "%global _use_internal_dependency_generator 0".
macros.kmp: Remove %__find_provides and %__find_requires.
* modalias.attr: New file.
* provided_ksyms.attr: Likewise.
* required_ksyms.attr: Likewise.
Resolves: #1942072
Resolves: #1942563
Signed-off-by: Eugene Syromiatnikov <esyr@redhat.com>
49 lines
1.5 KiB
Bash
Executable File
49 lines
1.5 KiB
Bash
Executable File
#! /bin/bash
|
|
|
|
IFS=$'\n'
|
|
|
|
for module in $(grep -E '/lib/modules/.+\.ko(\.gz|\.bz2|\.xz)?$') "$@"; do
|
|
tmpfile=""
|
|
if [ "x${module%.ko}" = "x${module}" ]; then
|
|
tmpfile=$(mktemp -t ${0##*/}.XXXXXX.ko)
|
|
proc_bin=
|
|
case "${module##*.}" in
|
|
xz)
|
|
proc_bin=xz
|
|
;;
|
|
bz2)
|
|
proc_bin=bzip2
|
|
;;
|
|
gz)
|
|
proc_bin=gzip
|
|
;;
|
|
esac
|
|
|
|
[ -n "$proc_bin" ] || continue
|
|
|
|
"$proc_bin" -d -c - < "$module" > "$tmpfile" || continue
|
|
module="$tmpfile"
|
|
fi
|
|
|
|
if [[ -n $(nm $module | sed -r -ne 's:^0*([0-9a-f]+) A __crc_(.+):0x\1 \2:p') ]]; then
|
|
nm $module \
|
|
| sed -r -ne 's:^0*([0-9a-f]+) A __crc_(.+):0x\1 \2:p' \
|
|
| awk --non-decimal-data '{printf("ksym(%s) = 0x%08x\n", $2, $1)}' \
|
|
| LC_ALL=C sort -u
|
|
else
|
|
ELFRODATA=$(readelf -R .rodata $module | awk '/0x/{printf $2$3$4$5}')
|
|
if [[ -n $(readelf -h $module | grep "little endian") ]]; then
|
|
RODATA=$(echo $ELFRODATA | sed 's/\(..\)\(..\)\(..\)\(..\)/\4\3\2\1/g')
|
|
else
|
|
RODATA=$ELFRODATA
|
|
fi
|
|
for sym in $(nm $module | sed -r -ne 's:^0*([0-9a-f]+) R __crc_(.+):0x\1 \2:p'); do
|
|
echo $sym $RODATA
|
|
done \
|
|
| awk --non-decimal-data '{printf("ksym(%s) = 0x%08s\n", $2, substr($3,($1*2)+1,8))}' \
|
|
| LC_ALL=C sort -u
|
|
fi
|
|
|
|
[ -z "$tmpfile" ] || rm -f -- "$tmpfile"
|
|
done
|