cache downloaded files(module) to cache/file/HASH
lib.sh download() is used by subfile handling in git.sh, e.g. crossgcc tarballs. they are not currently cached, but are downloaded directly in place. cache them, under cache/file/, saved with the name equal to the checksum, so: cache/file/CHECKSUM if the given cached file exists, use it as-is for simple copy, instead of curl. this avoids re-downloading a lot of crossgcc tarballs, where different coreboot trees may use some archives that are the same throughout. Signed-off-by: Leah Rowe <leah@libreboot.org>audit2
parent
0d55d7b23a
commit
8bb3730e7b
|
@ -1,5 +1,6 @@
|
|||
*~
|
||||
*.o
|
||||
/cache/
|
||||
/cbmk.err.log
|
||||
/repo/
|
||||
/docs/
|
||||
|
|
|
@ -180,18 +180,19 @@ singletree()
|
|||
|
||||
download()
|
||||
{
|
||||
dl_fail="y" # 1 url, 2 url backup, 3 destination, 4 checksum
|
||||
vendor_checksum "$4" "$3" 2>/dev/null || dl_fail="n"
|
||||
cached="cache/file/$4"
|
||||
dl_fail="n" # 1 url, 2 url backup, 3 destination, 4 checksum
|
||||
vendor_checksum "$4" "$cached" 2>/dev/null && dl_fail="y"
|
||||
[ "$dl_fail" = "n" ] && e "$3" f && return 0
|
||||
x_ mkdir -p "${3%/*}" && for url in "$1" "$2"; do
|
||||
x_ mkdir -p "${3%/*}" cache/file && for url in "$1" "$2"; do
|
||||
[ "$dl_fail" = "n" ] && break
|
||||
[ -z "$url" ] && continue
|
||||
x_ rm -f "$3"
|
||||
curl --location --retry 3 -A "$_ua" "$url" -o "$3" || \
|
||||
wget --tries 3 -U "$_ua" "$url" -O "$3" || continue
|
||||
vendor_checksum "$4" "$3" || dl_fail="n"
|
||||
done;
|
||||
[ "$dl_fail" = "y" ] && $err "$1 $2 $3 $4: not downloaded"; return 0
|
||||
x_ rm -f "$cached"
|
||||
curl --location --retry 3 -A "$_ua" "$url" -o "$cached" || \
|
||||
wget --tries 3 -U "$_ua" "$url" -O "$cached" || continue
|
||||
vendor_checksum "$4" "$cached" || dl_fail="n"
|
||||
done; [ "$dl_fail" = "y" ] && $err "$1 $2 $3 $4: not downloaded"
|
||||
[ "$cached" = "$3" ] || cp "$cached" "$3" || $err "!d cp $cached $3"; :
|
||||
}
|
||||
|
||||
vendor_checksum()
|
||||
|
|
Loading…
Reference in New Issue