Message ID | 20230425101216.1693998-1-felix.moessbauer@siemens.com |
---|---|
State | Accepted, archived |
Headers | show |
Series | [1/1] make compressed images better syncable | expand |
25/04/2023 13:12, 'Felix Moessbauer' via isar-users wrote: > For compression formats that support these options (currently only zstd), > we enable to include the sparseness information into the archive. By that, > disk IO on decompression of large zero blocks is speedup and storage usage > is decreased. Usually, this is handled automatically by the compression > tools, but not when sending output to stdout. > > In addition, we add the rsyncable flag to reduce the number > of changed blocks in the output when only changing some input blocks. > > Signed-off-by: Felix Moessbauer <felix.moessbauer@siemens.com> Applied to next, thanks.
diff --git a/meta/classes/imagetypes.bbclass b/meta/classes/imagetypes.bbclass index 2ee3240b..a80a6da5 100644 --- a/meta/classes/imagetypes.bbclass +++ b/meta/classes/imagetypes.bbclass @@ -101,5 +101,5 @@ CONVERSION_DEPS:gz = "gzip" CONVERSION_CMD:xz = "${SUDO_CHROOT} sh -c 'xz -c ${XZ_DEFAULTS} ${IMAGE_FILE_CHROOT} > ${IMAGE_FILE_CHROOT}.xz'" CONVERSION_DEPS:xz = "xz-utils" -CONVERSION_CMD:zst = "${SUDO_CHROOT} sh -c 'zstd -c ${ZSTD_DEFAULTS} ${IMAGE_FILE_CHROOT} > ${IMAGE_FILE_CHROOT}.zst'" +CONVERSION_CMD:zst = "${SUDO_CHROOT} sh -c 'zstd -c --sparse ${ZSTD_DEFAULTS} ${IMAGE_FILE_CHROOT} > ${IMAGE_FILE_CHROOT}.zst'" CONVERSION_DEPS:zst = "zstd" diff --git a/meta/conf/bitbake.conf b/meta/conf/bitbake.conf index f421050e..c660660a 100644 --- a/meta/conf/bitbake.conf +++ b/meta/conf/bitbake.conf @@ -141,7 +141,7 @@ XZ_DEFAULTS[vardepsexclude] += "XZ_MEMLIMIT XZ_THREADS" ZSTD_THREADS ?= "${@oe.utils.cpu_count(at_least=2)}" ZSTD_THREADS[vardepvalue] = "1" ZSTD_LEVEL ?= "19" -ZSTD_DEFAULTS ?= "-${ZSTD_LEVEL} --threads=${ZSTD_THREADS}" +ZSTD_DEFAULTS ?= "--rsyncable -${ZSTD_LEVEL} --threads=${ZSTD_THREADS}" ZSTD_DEFAULTS[vardepsexclude] += "ZSTD_LEVEL ZSTD_THREADS" BBINCLUDELOGS ??= "yes"
For compression formats that support these options (currently only zstd), we enable to include the sparseness information into the archive. By that, disk IO on decompression of large zero blocks is speedup and storage usage is decreased. Usually, this is handled automatically by the compression tools, but not when sending output to stdout. In addition, we add the rsyncable flag to reduce the number of changed blocks in the output when only changing some input blocks. Signed-off-by: Felix Moessbauer <felix.moessbauer@siemens.com> --- meta/classes/imagetypes.bbclass | 2 +- meta/conf/bitbake.conf | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-)