[PATCH v2 0/9] busybox: cleanup and fix config fragments

[PATCH v2 0/9] busybox: cleanup and fix config fragments

[Enrico Jörns]

Most of the busybox config fragments are useless since the respective options
were accidentally enabled in the 'defconfig' file during the defconfig update
to busybox 1.33.0 [1].

This is redundant and confusing and takes the ability to disable options by
removing the respective fragment file from the SRC_URI.

While cleaning this up, some minor inconsistencies were found and fixed.

The series starts with a defconfig sync to the current defaults to have a clear
base to work on.

The final .config for qemux86-64 is identical (empty diff) between a run on
master and a run based on these patches.

[1] 4335cd24 ("busybox: refresh the defconfig from 1.33.0")

-- 

v1 -> v2:

* rebase, regenerate, and adapt changes and commit messages to busybox 1.37
  * adapt to changes in sha1sum.cfg/sha256sum.cfg/sha_accel.cfg
  * adapt to dependency changes in CONFIG_HWCLOK/CONFIG_LONG_OPTS
* add missing deactivation of CONFIG_GETOPT in defconfig
* add missing deactivation of CONFIG_LOGREAD in defconfig
* reworked and fixed some long option issues
* added some required changes to existing fragments
* fix a few typos in commit messages

Enrico Jörns (9):
  busybox: sync config for busybox 1.37
  busybox: disable defconfig options from simpler fragments
  busybox: disable defconfig options from syslog.cfg
  busybox: mdev.cfg: remove CONFIG_SETSID
  busybox: disable defconfig options from login-utilities.cfg
  busybox: disable defconfig options from unicode.cfg
  busybox: disable defconfig options from getopts.cfg and move long opts
  busybox: disable long options in defconfig
  busybox: disable defconfig options from mount-via-label.cfg

 meta/recipes-core/busybox/busybox/defconfig   | 121 +++++++++---------
 meta/recipes-core/busybox/busybox/getopts.cfg |   2 -
 .../busybox/busybox/login-utilities.cfg       |   3 +
 .../recipes-core/busybox/busybox/longopts.cfg |   1 +
 meta/recipes-core/busybox/busybox/mdev.cfg    |   1 -
 .../busybox/busybox/mount-via-label.cfg       |   1 +
 6 files changed, 65 insertions(+), 64 deletions(-)

-- 
2.39.5

[PATCH v2 1/9] busybox: sync config for busybox 1.37

[Enrico Jörns]

Done by disabling all current cfg fragments in SRC_URI, running

  bitbake -c configure busybox

and copying the generated .config back to the defconfig file.
Changes to CONFIG_EXTRA_CFLAGS were omitted.

CONFIG_SHA1_HWACCEL and CONFIG_SHA256_HWACCEL are disabled explicitly in
the respective fragments (sha1sum.cfg, sha256sum.cfg), thus it makes
sense they've been enabled by default.

Signed-off-by: Enrico Jörns <ejo@pengutronix.de>
---
 meta/recipes-core/busybox/busybox/defconfig | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/meta/recipes-core/busybox/busybox/defconfig b/meta/recipes-core/busybox/busybox/defconfig
index d172ec46e9..6c5a8764ce 100644
--- a/meta/recipes-core/busybox/busybox/defconfig
+++ b/meta/recipes-core/busybox/busybox/defconfig
@@ -93,6 +93,9 @@ CONFIG_FEATURE_BUFFERS_USE_MALLOC=y
 # CONFIG_FEATURE_BUFFERS_GO_IN_BSS is not set
 CONFIG_PASSWORD_MINLEN=6
 CONFIG_MD5_SMALL=1
+CONFIG_SHA1_SMALL=3
+CONFIG_SHA1_HWACCEL=y
+CONFIG_SHA256_HWACCEL=y
 CONFIG_SHA3_SMALL=1
 CONFIG_FEATURE_NON_POSIX_CP=y
 # CONFIG_FEATURE_VERBOSE_CP_MESSAGE is not set
@@ -884,6 +887,7 @@ CONFIG_FEATURE_HWIB=y
 CONFIG_HOSTNAME=y
 CONFIG_DNSDOMAINNAME=y
 # CONFIG_HTTPD is not set
+CONFIG_FEATURE_HTTPD_PORT_DEFAULT=0
 # CONFIG_FEATURE_HTTPD_RANGES is not set
 # CONFIG_FEATURE_HTTPD_SETUID is not set
 # CONFIG_FEATURE_HTTPD_BASIC_AUTH is not set
@@ -977,6 +981,7 @@ CONFIG_FEATURE_TELNET_AUTOLOGIN=y
 CONFIG_FEATURE_TELNET_WIDTH=y
 # CONFIG_TELNETD is not set
 # CONFIG_FEATURE_TELNETD_STANDALONE is not set
+CONFIG_FEATURE_TELNETD_PORT_DEFAULT=0
 # CONFIG_FEATURE_TELNETD_INETD_WAIT is not set
 CONFIG_TFTP=y
 # CONFIG_FEATURE_TFTP_PROGRESS_BAR is not set
@@ -1015,6 +1020,7 @@ CONFIG_UDHCPC=y
 CONFIG_FEATURE_UDHCPC_ARPING=y
 CONFIG_FEATURE_UDHCPC_SANITIZEOPT=y
 CONFIG_UDHCPC_DEFAULT_SCRIPT="/usr/share/udhcpc/default.script"
+CONFIG_UDHCPC6_DEFAULT_SCRIPT=""
 # CONFIG_UDHCPC6 is not set
 # CONFIG_FEATURE_UDHCPC6_RFC3646 is not set
 # CONFIG_FEATURE_UDHCPC6_RFC4704 is not set
@@ -1024,6 +1030,7 @@ CONFIG_UDHCPC_DEFAULT_SCRIPT="/usr/share/udhcpc/default.script"
 #
 # Common options for DHCP applets
 #
+CONFIG_UDHCPC_DEFAULT_INTERFACE="eth0"
 # CONFIG_FEATURE_UDHCP_PORT is not set
 CONFIG_UDHCP_DEBUG=0
 CONFIG_UDHCPC_SLACK_FOR_BUGGY_SERVERS=80
-- 
2.39.5

[PATCH v2 2/9] busybox: disable defconfig options from simpler fragments

[Enrico Jörns]

This disables options from the defconfig file that are enabled by
explicit config fragments.

Having them enabled in the defconfig renders the fragments useless and
takes away the ability to disable options with

  SRC_URI:remove = "<fragment>.cfg".

The respective options were all deactivated once but got accidentally
enabled in 4335cd24 ("busybox: refresh the defconfig from 1.33.0").

This commit disables the features for:

- sha1sum.cfg
- sha256sum.cfg
- resize.cfg
- pgrep.cfg
- rev.cfg

Signed-off-by: Enrico Jörns <ejo@pengutronix.de>
---
 meta/recipes-core/busybox/busybox/defconfig | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/meta/recipes-core/busybox/busybox/defconfig b/meta/recipes-core/busybox/busybox/defconfig
index 6c5a8764ce..0ec65c65af 100644
--- a/meta/recipes-core/busybox/busybox/defconfig
+++ b/meta/recipes-core/busybox/busybox/defconfig
@@ -281,8 +281,8 @@ CONFIG_FEATURE_LS_USERNAME=y
 CONFIG_FEATURE_LS_COLOR=y
 # CONFIG_FEATURE_LS_COLOR_IS_DEFAULT is not set
 CONFIG_MD5SUM=y
-CONFIG_SHA1SUM=y
-CONFIG_SHA256SUM=y
+# CONFIG_SHA1SUM is not set
+# CONFIG_SHA256SUM is not set
 # CONFIG_SHA512SUM is not set
 # CONFIG_SHA3SUM is not set
 
@@ -386,8 +386,8 @@ CONFIG_DEFAULT_SETFONT_DIR=""
 CONFIG_LOADKMAP=y
 CONFIG_OPENVT=y
 CONFIG_RESET=y
-CONFIG_RESIZE=y
-CONFIG_FEATURE_RESIZE_PRINT=y
+# CONFIG_RESIZE is not set
+# CONFIG_FEATURE_RESIZE_PRINT is not set
 CONFIG_SETCONSOLE=y
 # CONFIG_FEATURE_SETCONSOLE_LONG_OPTIONS is not set
 # CONFIG_SETKEYCODES is not set
@@ -689,7 +689,7 @@ CONFIG_RDATE=y
 # CONFIG_RDEV is not set
 # CONFIG_READPROFILE is not set
 CONFIG_RENICE=y
-CONFIG_REV=y
+# CONFIG_REV is not set
 # CONFIG_RTCWAKE is not set
 # CONFIG_SCRIPT is not set
 # CONFIG_SCRIPTREPLAY is not set
@@ -1070,7 +1070,7 @@ CONFIG_KILLALL=y
 # CONFIG_LSOF is not set
 # CONFIG_MPSTAT is not set
 # CONFIG_NMETER is not set
-CONFIG_PGREP=y
+# CONFIG_PGREP is not set
 # CONFIG_PKILL is not set
 CONFIG_PIDOF=y
 # CONFIG_FEATURE_PIDOF_SINGLE is not set
-- 
2.39.5

[PATCH v2 3/9] busybox: disable defconfig options from syslog.cfg

[Enrico Jörns]

Accidentally enabled in
4335cd24 ("busybox: refresh the defconfig from 1.33.0").

Signed-off-by: Enrico Jörns <ejo@pengutronix.de>
---
 meta/recipes-core/busybox/busybox/defconfig | 26 +++++++++------------
 1 file changed, 11 insertions(+), 15 deletions(-)

diff --git a/meta/recipes-core/busybox/busybox/defconfig b/meta/recipes-core/busybox/busybox/defconfig
index 0ec65c65af..d2a210d535 100644
--- a/meta/recipes-core/busybox/busybox/defconfig
+++ b/meta/recipes-core/busybox/busybox/defconfig
@@ -1214,21 +1214,17 @@ CONFIG_FEATURE_SH_HISTFILESIZE=y
 # System Logging Utilities
 #
 CONFIG_KLOGD=y
-
-#
-# klogd should not be used together with syslog to kernel printk buffer
-#
 CONFIG_FEATURE_KLOGD_KLOGCTL=y
 CONFIG_LOGGER=y
-CONFIG_LOGREAD=y
-CONFIG_FEATURE_LOGREAD_REDUCED_LOCKING=y
-CONFIG_SYSLOGD=y
-CONFIG_FEATURE_ROTATE_LOGFILE=y
-CONFIG_FEATURE_REMOTE_LOG=y
-CONFIG_FEATURE_SYSLOGD_DUP=y
-CONFIG_FEATURE_SYSLOGD_CFG=y
+# CONFIG_LOGREAD is not set
+# CONFIG_FEATURE_LOGREAD_REDUCED_LOCKING is not set
+# CONFIG_SYSLOGD is not set
+# CONFIG_FEATURE_ROTATE_LOGFILE is not set
+# CONFIG_FEATURE_REMOTE_LOG is not set
+# CONFIG_FEATURE_SYSLOGD_DUP is not set
+# CONFIG_FEATURE_SYSLOGD_CFG is not set
 # CONFIG_FEATURE_SYSLOGD_PRECISE_TIMESTAMPS is not set
-CONFIG_FEATURE_SYSLOGD_READ_BUFFER_SIZE=256
-CONFIG_FEATURE_IPC_SYSLOG=y
-CONFIG_FEATURE_IPC_SYSLOG_BUFFER_SIZE=64
-CONFIG_FEATURE_KMSG_SYSLOG=y
+CONFIG_FEATURE_SYSLOGD_READ_BUFFER_SIZE=0
+# CONFIG_FEATURE_IPC_SYSLOG is not set
+CONFIG_FEATURE_IPC_SYSLOG_BUFFER_SIZE=0
+# CONFIG_FEATURE_KMSG_SYSLOG is not set
-- 
2.39.5

[PATCH v2 4/9] busybox: mdev.cfg: remove CONFIG_SETSID

[Enrico Jörns]

Was explicitly enabled in defconfig in
456010a0 ("busybox: Enable mountpoint and setsid applets").

Signed-off-by: Enrico Jörns <ejo@pengutronix.de>
---
 meta/recipes-core/busybox/busybox/mdev.cfg | 1 -
 1 file changed, 1 deletion(-)

diff --git a/meta/recipes-core/busybox/busybox/mdev.cfg b/meta/recipes-core/busybox/busybox/mdev.cfg
index 143e6097cb..54d23597cc 100644
--- a/meta/recipes-core/busybox/busybox/mdev.cfg
+++ b/meta/recipes-core/busybox/busybox/mdev.cfg
@@ -5,7 +5,6 @@ CONFIG_FEATURE_MDEV_RENAME_REGEXP=y
 CONFIG_FEATURE_MDEV_EXEC=y
 CONFIG_FEATURE_MDEV_LOAD_FIRMWARE=y
 
-CONFIG_SETSID=y
 CONFIG_CTTYHACK=y
 
 CONFIG_FEATURE_SHADOWPASSWDS=y
-- 
2.39.5

[PATCH v2 5/9] busybox: disable defconfig options from login-utilities.cfg

[Enrico Jörns]

Accidentally enabled in
4335cd24 ("busybox: refresh the defconfig from 1.33.0").

Since this would also enforce wrong ID ranges due to busbox kconfig's
bad handling of 'inactive' options, manually add them to the
login-utilities.cfg fragment.

Signed-off-by: Enrico Jörns <ejo@pengutronix.de>
---
 meta/recipes-core/busybox/busybox/defconfig   | 28 +++++++++----------
 .../busybox/busybox/login-utilities.cfg       |  3 ++
 2 files changed, 17 insertions(+), 14 deletions(-)

diff --git a/meta/recipes-core/busybox/busybox/defconfig b/meta/recipes-core/busybox/busybox/defconfig
index d2a210d535..116286227b 100644
--- a/meta/recipes-core/busybox/busybox/defconfig
+++ b/meta/recipes-core/busybox/busybox/defconfig
@@ -524,41 +524,41 @@ CONFIG_INIT_TERMINAL_TYPE=""
 #
 # Login/Password Management Utilities
 #
-CONFIG_FEATURE_SHADOWPASSWDS=y
+# CONFIG_FEATURE_SHADOWPASSWDS is not set
 # CONFIG_USE_BB_PWD_GRP is not set
 # CONFIG_USE_BB_SHADOW is not set
 CONFIG_USE_BB_CRYPT=y
 # CONFIG_USE_BB_CRYPT_SHA is not set
 # CONFIG_ADD_SHELL is not set
 # CONFIG_REMOVE_SHELL is not set
-CONFIG_ADDGROUP=y
+# CONFIG_ADDGROUP is not set
 # CONFIG_FEATURE_ADDUSER_TO_GROUP is not set
-CONFIG_ADDUSER=y
+# CONFIG_ADDUSER is not set
 # CONFIG_FEATURE_CHECK_NAMES is not set
-CONFIG_LAST_ID=60000
-CONFIG_FIRST_SYSTEM_ID=100
-CONFIG_LAST_SYSTEM_ID=999
+CONFIG_LAST_ID=0
+CONFIG_FIRST_SYSTEM_ID=0
+CONFIG_LAST_SYSTEM_ID=0
 # CONFIG_CHPASSWD is not set
 CONFIG_FEATURE_DEFAULT_PASSWD_ALGO=""
 # CONFIG_CRYPTPW is not set
 # CONFIG_MKPASSWD is not set
-CONFIG_DELUSER=y
-CONFIG_DELGROUP=y
+# CONFIG_DELUSER is not set
+# CONFIG_DELGROUP is not set
 # CONFIG_FEATURE_DEL_USER_FROM_GROUP is not set
-CONFIG_GETTY=y
-CONFIG_LOGIN=y
+# CONFIG_GETTY is not set
+# CONFIG_LOGIN is not set
 # CONFIG_LOGIN_SESSION_AS_CHILD is not set
 # CONFIG_LOGIN_SCRIPTS is not set
 # CONFIG_FEATURE_NOLOGIN is not set
 # CONFIG_FEATURE_SECURETTY is not set
-CONFIG_PASSWD=y
+# CONFIG_PASSWD is not set
 # CONFIG_FEATURE_PASSWD_WEAK_CHECK is not set
-CONFIG_SU=y
+# CONFIG_SU is not set
 # CONFIG_FEATURE_SU_SYSLOG is not set
 # CONFIG_FEATURE_SU_CHECKS_SHELLS is not set
 # CONFIG_FEATURE_SU_BLANK_PW_NEEDS_SECURE_TTY is not set
-CONFIG_SULOGIN=y
-CONFIG_VLOCK=y
+# CONFIG_SULOGIN is not set
+# CONFIG_VLOCK is not set
 
 #
 # Linux Ext2 FS Progs
diff --git a/meta/recipes-core/busybox/busybox/login-utilities.cfg b/meta/recipes-core/busybox/busybox/login-utilities.cfg
index cc9b2db502..8a21e80619 100644
--- a/meta/recipes-core/busybox/busybox/login-utilities.cfg
+++ b/meta/recipes-core/busybox/busybox/login-utilities.cfg
@@ -1,5 +1,8 @@
 CONFIG_FEATURE_SHADOWPASSWDS=y
 CONFIG_ADDUSER=y
+CONFIG_LAST_ID=60000
+CONFIG_FIRST_SYSTEM_ID=100
+CONFIG_LAST_SYSTEM_ID=999
 CONFIG_FEATURE_ADDUSER_LONG_OPTIONS=y
 CONFIG_ADDGROUP=y
 CONFIG_FEATURE_ADDGROUP_LONG_OPTIONS=y
-- 
2.39.5

[PATCH v2 6/9] busybox: disable defconfig options from unicode.cfg

[Enrico Jörns]

Accidentally enabled in
4335cd24 ("busybox: refresh the defconfig from 1.33.0").

Signed-off-by: Enrico Jörns <ejo@pengutronix.de>
---
 meta/recipes-core/busybox/busybox/defconfig | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/meta/recipes-core/busybox/busybox/defconfig b/meta/recipes-core/busybox/busybox/defconfig
index 116286227b..a65ae674d8 100644
--- a/meta/recipes-core/busybox/busybox/defconfig
+++ b/meta/recipes-core/busybox/busybox/defconfig
@@ -116,13 +116,13 @@ CONFIG_FEATURE_EDITING_FANCY_PROMPT=y
 # CONFIG_FEATURE_EDITING_WINCH is not set
 # CONFIG_FEATURE_EDITING_ASK_TERMINAL is not set
 # CONFIG_LOCALE_SUPPORT is not set
-CONFIG_UNICODE_SUPPORT=y
+# CONFIG_UNICODE_SUPPORT is not set
 # CONFIG_UNICODE_USING_LOCALE is not set
-CONFIG_FEATURE_CHECK_UNICODE_IN_ENV=y
-CONFIG_SUBST_WCHAR=63
-CONFIG_LAST_SUPPORTED_WCHAR=767
-CONFIG_UNICODE_COMBINING_WCHARS=y
-CONFIG_UNICODE_WIDE_WCHARS=y
+# CONFIG_FEATURE_CHECK_UNICODE_IN_ENV is not set
+CONFIG_SUBST_WCHAR=0
+CONFIG_LAST_SUPPORTED_WCHAR=0
+# CONFIG_UNICODE_COMBINING_WCHARS is not set
+# CONFIG_UNICODE_WIDE_WCHARS is not set
 # CONFIG_UNICODE_BIDI_SUPPORT is not set
 # CONFIG_UNICODE_NEUTRAL_TABLE is not set
 # CONFIG_UNICODE_PRESERVE_BROKEN is not set
-- 
2.39.5

[PATCH v2 7/9] busybox: disable defconfig options from getopts.cfg and move long opts

[Enrico Jörns]

This disables getopts.cfg fragment options in the defconfig which were
accidentally enabled in 4335cd24 ("busybox: refresh the defconfig from
1.33.0").

Since there is longopts.cfg for long options, move
CONFIG_FEATURE_GETOPT_LONG there. It was probably not placed there
originally since it does not match the naming scheme of the other long
options.

Note that CONFIG_LONG_OPTS will just be dropped since it is enabled in
defconfig anyway.

Signed-off-by: Enrico Jörns <ejo@pengutronix.de>
---
 meta/recipes-core/busybox/busybox/defconfig    | 4 ++--
 meta/recipes-core/busybox/busybox/getopts.cfg  | 2 --
 meta/recipes-core/busybox/busybox/longopts.cfg | 1 +
 3 files changed, 3 insertions(+), 4 deletions(-)

diff --git a/meta/recipes-core/busybox/busybox/defconfig b/meta/recipes-core/busybox/busybox/defconfig
index a65ae674d8..7481ea2d6b 100644
--- a/meta/recipes-core/busybox/busybox/defconfig
+++ b/meta/recipes-core/busybox/busybox/defconfig
@@ -636,8 +636,8 @@ CONFIG_FLOCK=y
 # CONFIG_FSCK_MINIX is not set
 # CONFIG_FSFREEZE is not set
 CONFIG_FSTRIM=y
-CONFIG_GETOPT=y
-CONFIG_FEATURE_GETOPT_LONG=y
+# CONFIG_GETOPT is not set
+# CONFIG_FEATURE_GETOPT_LONG is not set
 CONFIG_HEXDUMP=y
 # CONFIG_HD is not set
 # CONFIG_XXD is not set
diff --git a/meta/recipes-core/busybox/busybox/getopts.cfg b/meta/recipes-core/busybox/busybox/getopts.cfg
index 8db0a4a8f2..26a2e7add9 100644
--- a/meta/recipes-core/busybox/busybox/getopts.cfg
+++ b/meta/recipes-core/busybox/busybox/getopts.cfg
@@ -1,3 +1 @@
 CONFIG_GETOPT=y
-CONFIG_LONG_OPTS=y
-CONFIG_FEATURE_GETOPT_LONG=y
diff --git a/meta/recipes-core/busybox/busybox/longopts.cfg b/meta/recipes-core/busybox/busybox/longopts.cfg
index dcfab99919..3255608f64 100644
--- a/meta/recipes-core/busybox/busybox/longopts.cfg
+++ b/meta/recipes-core/busybox/busybox/longopts.cfg
@@ -13,3 +13,4 @@ CONFIG_FEATURE_FTPGETPUT_LONG_OPTIONS=y
 CONFIG_FEATURE_IPCALC_LONG_OPTIONS=y
 CONFIG_FEATURE_NSLOOKUP_LONG_OPTIONS=y
 CONFIG_FEATURE_WGET_LONG_OPTIONS=y
+CONFIG_FEATURE_GETOPT_LONG=y
-- 
2.39.5

[PATCH v2 8/9] busybox: disable long options in defconfig

[Enrico Jörns]

These are activated by the longopts.cfg fragment, thus there should be
no need to set them in the defconfig.
The individual long options where enabled without mentioning a specific
need, mostly while adding the corresponding command, thus it should be
safe to disable them.

Note that we cannot disable the general 'LONG_OPTS' feature in the
defconfig since this is e.g. required by CONFIG_HWCLOCK (since v1.37).

Signed-off-by: Enrico Jörns <ejo@pengutronix.de>
---
 meta/recipes-core/busybox/busybox/defconfig | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/meta/recipes-core/busybox/busybox/defconfig b/meta/recipes-core/busybox/busybox/defconfig
index 7481ea2d6b..ab74d2616d 100644
--- a/meta/recipes-core/busybox/busybox/defconfig
+++ b/meta/recipes-core/busybox/busybox/defconfig
@@ -148,7 +148,7 @@ CONFIG_FEATURE_SEAMLESS_Z=y
 # CONFIG_UNCOMPRESS is not set
 CONFIG_GUNZIP=y
 CONFIG_ZCAT=y
-CONFIG_FEATURE_GUNZIP_LONG_OPTIONS=y
+# CONFIG_FEATURE_GUNZIP_LONG_OPTIONS is not set
 CONFIG_BUNZIP2=y
 CONFIG_BZCAT=y
 # CONFIG_UNLZMA is not set
@@ -399,10 +399,10 @@ CONFIG_SETCONSOLE=y
 #
 # CONFIG_PIPE_PROGRESS is not set
 CONFIG_RUN_PARTS=y
-CONFIG_FEATURE_RUN_PARTS_LONG_OPTIONS=y
+# CONFIG_FEATURE_RUN_PARTS_LONG_OPTIONS is not set
 # CONFIG_FEATURE_RUN_PARTS_FANCY is not set
 CONFIG_START_STOP_DAEMON=y
-CONFIG_FEATURE_START_STOP_DAEMON_LONG_OPTIONS=y
+# CONFIG_FEATURE_START_STOP_DAEMON_LONG_OPTIONS is not set
 CONFIG_FEATURE_START_STOP_DAEMON_FANCY=y
 CONFIG_WHICH=y
 
@@ -1000,7 +1000,7 @@ CONFIG_TRACEROUTE=y
 # CONFIG_FEATURE_TUNCTL_UG is not set
 # CONFIG_VCONFIG is not set
 CONFIG_WGET=y
-CONFIG_FEATURE_WGET_LONG_OPTIONS=y
+# CONFIG_FEATURE_WGET_LONG_OPTIONS is not set
 CONFIG_FEATURE_WGET_STATUSBAR=y
 CONFIG_FEATURE_WGET_FTP=y
 CONFIG_FEATURE_WGET_AUTHENTICATION=y
-- 
2.39.5

[PATCH v2 9/9] busybox: disable defconfig options from mount-via-label.cfg

[Enrico Jörns]

Accidentally enabled in
6a6c6442 ("busybox: refresh the defconfig from 1.33.0").

With this, also CONFIG_FEATURE_VOLUMEID_EROFS was enabled accidentally.
To not break possible existing use cases, we add it to the existing
fragment manually.

Signed-off-by: Enrico Jörns <ejo@pengutronix.de>
---
 meta/recipes-core/busybox/busybox/defconfig   | 24 ++++++++-----------
 .../busybox/busybox/mount-via-label.cfg       |  1 +
 2 files changed, 11 insertions(+), 14 deletions(-)

diff --git a/meta/recipes-core/busybox/busybox/defconfig b/meta/recipes-core/busybox/busybox/defconfig
index ab74d2616d..38b82457b7 100644
--- a/meta/recipes-core/busybox/busybox/defconfig
+++ b/meta/recipes-core/busybox/busybox/defconfig
@@ -605,8 +605,8 @@ CONFIG_DEFAULT_DEPMOD_FILE="modules.dep"
 # CONFIG_ACPID is not set
 # CONFIG_FEATURE_ACPID_COMPAT is not set
 # CONFIG_BLKDISCARD is not set
-CONFIG_BLKID=y
-CONFIG_FEATURE_BLKID_TYPE=y
+# CONFIG_BLKID is not set
+# CONFIG_FEATURE_BLKID_TYPE is not set
 # CONFIG_BLOCKDEV is not set
 # CONFIG_CAL is not set
 # CONFIG_CHRT is not set
@@ -674,7 +674,7 @@ CONFIG_MOUNT=y
 # CONFIG_FEATURE_MOUNT_FAKE is not set
 # CONFIG_FEATURE_MOUNT_VERBOSE is not set
 # CONFIG_FEATURE_MOUNT_HELPERS is not set
-CONFIG_FEATURE_MOUNT_LABEL=y
+# CONFIG_FEATURE_MOUNT_LABEL is not set
 # CONFIG_FEATURE_MOUNT_NFS is not set
 # CONFIG_FEATURE_MOUNT_CIFS is not set
 CONFIG_FEATURE_MOUNT_FLAGS=y
@@ -722,19 +722,15 @@ CONFIG_FEATURE_UMOUNT_ALL=y
 CONFIG_FEATURE_MOUNT_LOOP=y
 CONFIG_FEATURE_MOUNT_LOOP_CREATE=y
 # CONFIG_FEATURE_MTAB_SUPPORT is not set
-CONFIG_VOLUMEID=y
-
-#
-# Filesystem/Volume identification
-#
+# CONFIG_VOLUMEID is not set
 # CONFIG_FEATURE_VOLUMEID_BCACHE is not set
-CONFIG_FEATURE_VOLUMEID_BTRFS=y
+# CONFIG_FEATURE_VOLUMEID_BTRFS is not set
 # CONFIG_FEATURE_VOLUMEID_CRAMFS is not set
-CONFIG_FEATURE_VOLUMEID_EROFS=y
+# CONFIG_FEATURE_VOLUMEID_EROFS is not set
 # CONFIG_FEATURE_VOLUMEID_EXFAT is not set
-CONFIG_FEATURE_VOLUMEID_EXT=y
-CONFIG_FEATURE_VOLUMEID_F2FS=y
-CONFIG_FEATURE_VOLUMEID_FAT=y
+# CONFIG_FEATURE_VOLUMEID_EXT is not set
+# CONFIG_FEATURE_VOLUMEID_F2FS is not set
+# CONFIG_FEATURE_VOLUMEID_FAT is not set
 # CONFIG_FEATURE_VOLUMEID_HFS is not set
 # CONFIG_FEATURE_VOLUMEID_ISO9660 is not set
 # CONFIG_FEATURE_VOLUMEID_JFS is not set
@@ -748,7 +744,7 @@ CONFIG_FEATURE_VOLUMEID_FAT=y
 # CONFIG_FEATURE_VOLUMEID_OCFS2 is not set
 # CONFIG_FEATURE_VOLUMEID_REISERFS is not set
 # CONFIG_FEATURE_VOLUMEID_ROMFS is not set
-CONFIG_FEATURE_VOLUMEID_SQUASHFS=y
+# CONFIG_FEATURE_VOLUMEID_SQUASHFS is not set
 # CONFIG_FEATURE_VOLUMEID_SYSV is not set
 # CONFIG_FEATURE_VOLUMEID_UBIFS is not set
 # CONFIG_FEATURE_VOLUMEID_UDF is not set
diff --git a/meta/recipes-core/busybox/busybox/mount-via-label.cfg b/meta/recipes-core/busybox/busybox/mount-via-label.cfg
index 5a285de7cd..c9ae95c8c2 100644
--- a/meta/recipes-core/busybox/busybox/mount-via-label.cfg
+++ b/meta/recipes-core/busybox/busybox/mount-via-label.cfg
@@ -4,6 +4,7 @@ CONFIG_FEATURE_MOUNT_LABEL=y
 CONFIG_FEATURE_SWAPONOFF_LABEL=y
 CONFIG_VOLUMEID=y
 CONFIG_FEATURE_VOLUMEID_BTRFS=y
+CONFIG_FEATURE_VOLUMEID_EROFS=y
 CONFIG_FEATURE_VOLUMEID_EXT=y
 CONFIG_FEATURE_VOLUMEID_F2FS=y
 CONFIG_FEATURE_VOLUMEID_FAT=y
-- 
2.39.5

Re: [PATCH v2 2/9] busybox: disable defconfig options from simpler fragments

[Andrej Valek]

[-- Attachment #1: Type: text/plain, Size: 4399 bytes --]

Hello Enrico,

I tried to follow (and applied) your patch, but it looks little bit 
inconsistent to me. In the first round you enabled SHAx_HWACCEL, which 
is broken on some architectures. That's the reason, why is disabled by 
default. Ok, so first apply looks like this:

  # CONFIG_FEATURE_BUFFERS_GO_IN_BSS is not set
  CONFIG_PASSWORD_MINLEN=6
  CONFIG_MD5_SMALL=1
+CONFIG_SHA1_SMALL=3
+CONFIG_SHA1_HWACCEL=y
+CONFIG_SHA256_HWACCEL=y
  CONFIG_SHA3_SMALL=1
  CONFIG_FEATURE_NON_POSIX_CP=y
  # CONFIG_FEATURE_VERBOSE_CP_MESSAGE is not set
@@ -884,6 +887,7 @@ CONFIG_FEATURE_HWIB=y

Then I applied next patch:

  # CONFIG_FEATURE_BUFFERS_GO_IN_BSS is not set
  CONFIG_PASSWORD_MINLEN=6
  CONFIG_MD5_SMALL=1
+CONFIG_SHA1_SMALL=3
+CONFIG_SHA1_HWACCEL=y
+CONFIG_SHA256_HWACCEL=y
  CONFIG_SHA3_SMALL=1
  CONFIG_FEATURE_NON_POSIX_CP=y
  # CONFIG_FEATURE_VERBOSE_CP_MESSAGE is not set
@@ -278,8 +281,8 @@ CONFIG_FEATURE_LS_USERNAME=y
  CONFIG_FEATURE_LS_COLOR=y
  # CONFIG_FEATURE_LS_COLOR_IS_DEFAULT is not set
  CONFIG_MD5SUM=y
-CONFIG_SHA1SUM=y
-CONFIG_SHA256SUM=y
+# CONFIG_SHA1SUM is not set
+# CONFIG_SHA256SUM is not set
  # CONFIG_SHA512SUM is not set
  # CONFIG_SHA3SUM is not set

@@ -383,8 +386,8 @@ CONFIG_DEFAULT_SETFONT_DIR=""

and you disabled CONFIG_SHA1SUM, but keep enabled the acceleration? Is 
there any special reason for that?

Anyway, there was one commit in the usptream which was touching the 
acceleration. So I have to try it on the broken machine and see if it's 
still broken or not.

> On Fri, 2024-10-25 at 19:26 +0200, Andrej wrote:
>> Ok, I was thinking that it was done by intention, but it happened 14y
>> ago
>> https://github.com/openembedded/openembedded-core/commit/615a98ed9a02
>> 1da245513790c064761a0a5a67e9 . So I guess not.
> I am not sure if I fully understand what you are referring to here.
> Could you provide some details about what happened there?
> OE didn't have those fragments 14y ago I guess.
So the idea behind is that you're going to remove the fragments and keep 
just a one defconfig. But according to your commits you kept the 
fragments there, even if you put some values into defconfig.

Regards,
Andrej


On 28.10.2024 09:29, Enrico Jörns wrote:
> This disables options from the defconfig file that are enabled by
> explicit config fragments.
>
> Having them enabled in the defconfig renders the fragments useless and
> takes away the ability to disable options with
>
>    SRC_URI:remove = "<fragment>.cfg".
>
> The respective options were all deactivated once but got accidentally
> enabled in 4335cd24 ("busybox: refresh the defconfig from 1.33.0").
>
> This commit disables the features for:
>
> - sha1sum.cfg
> - sha256sum.cfg
> - resize.cfg
> - pgrep.cfg
> - rev.cfg
>
> Signed-off-by: Enrico Jörns<ejo@pengutronix.de>
> ---
>   meta/recipes-core/busybox/busybox/defconfig | 12 ++++++------
>   1 file changed, 6 insertions(+), 6 deletions(-)
>
> diff --git a/meta/recipes-core/busybox/busybox/defconfig b/meta/recipes-core/busybox/busybox/defconfig
> index 6c5a8764ce..0ec65c65af 100644
> --- a/meta/recipes-core/busybox/busybox/defconfig
> +++ b/meta/recipes-core/busybox/busybox/defconfig
> @@ -281,8 +281,8 @@ CONFIG_FEATURE_LS_USERNAME=y
>   CONFIG_FEATURE_LS_COLOR=y
>   # CONFIG_FEATURE_LS_COLOR_IS_DEFAULT is not set
>   CONFIG_MD5SUM=y
> -CONFIG_SHA1SUM=y
> -CONFIG_SHA256SUM=y
> +# CONFIG_SHA1SUM is not set
> +# CONFIG_SHA256SUM is not set
>   # CONFIG_SHA512SUM is not set
>   # CONFIG_SHA3SUM is not set
>   
> @@ -386,8 +386,8 @@ CONFIG_DEFAULT_SETFONT_DIR=""
>   CONFIG_LOADKMAP=y
>   CONFIG_OPENVT=y
>   CONFIG_RESET=y
> -CONFIG_RESIZE=y
> -CONFIG_FEATURE_RESIZE_PRINT=y
> +# CONFIG_RESIZE is not set
> +# CONFIG_FEATURE_RESIZE_PRINT is not set
>   CONFIG_SETCONSOLE=y
>   # CONFIG_FEATURE_SETCONSOLE_LONG_OPTIONS is not set
>   # CONFIG_SETKEYCODES is not set
> @@ -689,7 +689,7 @@ CONFIG_RDATE=y
>   # CONFIG_RDEV is not set
>   # CONFIG_READPROFILE is not set
>   CONFIG_RENICE=y
> -CONFIG_REV=y
> +# CONFIG_REV is not set
>   # CONFIG_RTCWAKE is not set
>   # CONFIG_SCRIPT is not set
>   # CONFIG_SCRIPTREPLAY is not set
> @@ -1070,7 +1070,7 @@ CONFIG_KILLALL=y
>   # CONFIG_LSOF is not set
>   # CONFIG_MPSTAT is not set
>   # CONFIG_NMETER is not set
> -CONFIG_PGREP=y
> +# CONFIG_PGREP is not set
>   # CONFIG_PKILL is not set
>   CONFIG_PIDOF=y
>   # CONFIG_FEATURE_PIDOF_SINGLE is not set

[-- Attachment #2: Type: text/html, Size: 5861 bytes --]

Re: [PATCH v2 2/9] busybox: disable defconfig options from simpler fragments

[Enrico Jörns]

Hej Andrej,

Am Montag, dem 28.10.2024 um 10:06 +0100 schrieb Andrej Valek:
>  Hello Enrico,
>  
>  I tried to follow (and applied) your patch, but it looks little bit inconsistent to me. In the
> first round you enabled SHAx_HWACCEL, which is broken on some architectures. That's the reason,
> why is disabled by default. Ok, so first apply looks like this:

let me rephrase my commit message.

The SHAx_HWACCEL is broken on some architectures, right.
However, when you take the defconfig, without the fragments, it will currently be enabled.
Regardless of what's in my patch.

Why? Because CONFIG_SHA1_HWACCEL in Busybox' kconfig is 'default y', has no dependency on other
options and is not listed in the 'defconfig' file so far.

Thus, my first patch should not change the current situation.

>   # CONFIG_FEATURE_BUFFERS_GO_IN_BSS is not set
>   CONFIG_PASSWORD_MINLEN=6
>   CONFIG_MD5_SMALL=1
>  +CONFIG_SHA1_SMALL=3
>  +CONFIG_SHA1_HWACCEL=y
>  +CONFIG_SHA256_HWACCEL=y
>   CONFIG_SHA3_SMALL=1
>   CONFIG_FEATURE_NON_POSIX_CP=y
>   # CONFIG_FEATURE_VERBOSE_CP_MESSAGE is not set
>  @@ -884,6 +887,7 @@ CONFIG_FEATURE_HWIB=y
>  
>  Then I applied next patch:
>  
>   # CONFIG_FEATURE_BUFFERS_GO_IN_BSS is not set
>   CONFIG_PASSWORD_MINLEN=6
>   CONFIG_MD5_SMALL=1
>  +CONFIG_SHA1_SMALL=3
>  +CONFIG_SHA1_HWACCEL=y
>  +CONFIG_SHA256_HWACCEL=y
>   CONFIG_SHA3_SMALL=1
>   CONFIG_FEATURE_NON_POSIX_CP=y
>   # CONFIG_FEATURE_VERBOSE_CP_MESSAGE is not set
>  @@ -278,8 +281,8 @@ CONFIG_FEATURE_LS_USERNAME=y
>   CONFIG_FEATURE_LS_COLOR=y
>   # CONFIG_FEATURE_LS_COLOR_IS_DEFAULT is not set
>   CONFIG_MD5SUM=y
>  -CONFIG_SHA1SUM=y
>  -CONFIG_SHA256SUM=y
>  +# CONFIG_SHA1SUM is not set
>  +# CONFIG_SHA256SUM is not set
>   # CONFIG_SHA512SUM is not set
>   # CONFIG_SHA3SUM is not set
>   
>  @@ -383,8 +386,8 @@ CONFIG_DEFAULT_SETFONT_DIR=""
>  
>  and you disabled CONFIG_SHA1SUM, but keep enabled the acceleration? Is there any special reason
> for that?

There is one, because its the inverse of what the fragments apply, which is:

  $ cat ./meta/recipes-core/busybox/busybox/sha1sum.cfg                                            
  CONFIG_SHA1SUM=y
  CONFIG_SHA1_SMALL=3
  # CONFIG_SHA1_HWACCEL is not set

  $ cat ./meta/recipes-core/busybox/busybox/sha256sum.cfg 
  CONFIG_SHA256SUM=y
  # CONFIG_SHA256_HWACCEL is not set

I'm pretty sure not having touched the fragments.

The CONFIG_SHA1_HWACCEL options could have been changed in the defconfig, too.
But this is nothing that I've brought in. It's a result of
https://github.com/openembedded/openembedded-core/commit/21753f16a364e32050cf8d79bfa7e0f89be52ce7

Thus I'd assume my changes are consistent. Am I wrong?

>  Anyway, there was one commit in the usptream which was touching the acceleration. So I have to
> try it on the broken machine and see if it's still broken or not.

If you refer to oe-core upstream, yes there have been changes.
This is why I needed to rebase and rework my branch.
Or are there other changes I am not aware of, yet?

> 
> > On Fri, 2024-10-25 at 19:26 +0200, Andrej wrote:
> >  
> > > Ok, I was thinking that it was done by intention, but it happened 14y
> > > ago 
> > > https://github.com/openembedded/openembedded-core/commit/615a98ed9a02
> > > 1da245513790c064761a0a5a67e9 . So I guess not. 
>  
> > I am not sure if I fully understand what you are referring to here.
> > Could you provide some details about what happened there?
> > OE didn't have those fragments 14y ago I guess.
>  So the idea behind is that you're going to remove the fragments and keep just a one defconfig.
> But according to your commits you kept the fragments there, even if you put some values into
> defconfig.

Still not sure. I had no intention of removing the fragments.
And I'm unsure how this relates to the referenced commit.
My intention was to actually make them useful again, which is quite the opposite I guess.


Regards, Enrico

>  Regards,
>  Andrej
> 
>  
> On 28.10.2024 09:29, Enrico Jörns wrote:
>  
> > This disables options from the defconfig file that are enabled by
> > explicit config fragments.
> > 
> > Having them enabled in the defconfig renders the fragments useless and
> > takes away the ability to disable options with
> > 
> >   SRC_URI:remove = "<fragment>.cfg".
> > 
> > The respective options were all deactivated once but got accidentally
> > enabled in 4335cd24 ("busybox: refresh the defconfig from 1.33.0").
> > 
> > This commit disables the features for:
> > 
> > - sha1sum.cfg
> > - sha256sum.cfg
> > - resize.cfg
> > - pgrep.cfg
> > - rev.cfg
> > 
> > Signed-off-by: Enrico Jörns <ejo@pengutronix.de>
> > ---
> >  meta/recipes-core/busybox/busybox/defconfig | 12 ++++++------
> >  1 file changed, 6 insertions(+), 6 deletions(-)
> > 
> > diff --git a/meta/recipes-core/busybox/busybox/defconfig b/meta/recipes-
> > core/busybox/busybox/defconfig
> > index 6c5a8764ce..0ec65c65af 100644
> > --- a/meta/recipes-core/busybox/busybox/defconfig
> > +++ b/meta/recipes-core/busybox/busybox/defconfig
> > @@ -281,8 +281,8 @@ CONFIG_FEATURE_LS_USERNAME=y
> >  CONFIG_FEATURE_LS_COLOR=y
> >  # CONFIG_FEATURE_LS_COLOR_IS_DEFAULT is not set
> >  CONFIG_MD5SUM=y
> > -CONFIG_SHA1SUM=y
> > -CONFIG_SHA256SUM=y
> > +# CONFIG_SHA1SUM is not set
> > +# CONFIG_SHA256SUM is not set
> >  # CONFIG_SHA512SUM is not set
> >  # CONFIG_SHA3SUM is not set
> >  
> > @@ -386,8 +386,8 @@ CONFIG_DEFAULT_SETFONT_DIR=""
> >  CONFIG_LOADKMAP=y
> >  CONFIG_OPENVT=y
> >  CONFIG_RESET=y
> > -CONFIG_RESIZE=y
> > -CONFIG_FEATURE_RESIZE_PRINT=y
> > +# CONFIG_RESIZE is not set
> > +# CONFIG_FEATURE_RESIZE_PRINT is not set
> >  CONFIG_SETCONSOLE=y
> >  # CONFIG_FEATURE_SETCONSOLE_LONG_OPTIONS is not set
> >  # CONFIG_SETKEYCODES is not set
> > @@ -689,7 +689,7 @@ CONFIG_RDATE=y
> >  # CONFIG_RDEV is not set
> >  # CONFIG_READPROFILE is not set
> >  CONFIG_RENICE=y
> > -CONFIG_REV=y
> > +# CONFIG_REV is not set
> >  # CONFIG_RTCWAKE is not set
> >  # CONFIG_SCRIPT is not set
> >  # CONFIG_SCRIPTREPLAY is not set
> > @@ -1070,7 +1070,7 @@ CONFIG_KILLALL=y
> >  # CONFIG_LSOF is not set
> >  # CONFIG_MPSTAT is not set
> >  # CONFIG_NMETER is not set
> > -CONFIG_PGREP=y
> > +# CONFIG_PGREP is not set
> >  # CONFIG_PKILL is not set
> >  CONFIG_PIDOF=y
> >  # CONFIG_FEATURE_PIDOF_SINGLE is not set
>  
>  

-- 
Pengutronix e.K.                           | Enrico Jörns                |
Embedded Linux Consulting & Support        | https://www.pengutronix.de/ |
Steuerwalder Str. 21                       | Phone: +49-5121-206917-180  |
31137 Hildesheim, Germany                  | Fax:   +49-5121-206917-9    |

Re: [OE-core] [PATCH v2 2/9] busybox: disable defconfig options from simpler fragments

[Andrej Valek]

Hello Enrico,

On 28.10.2024 11:39, Enrico Jörns wrote:
> Hej Andrej,
>
> Am Montag, dem 28.10.2024 um 10:06 +0100 schrieb Andrej Valek:
>>   Hello Enrico,
>>   
>>   I tried to follow (and applied) your patch, but it looks little bit inconsistent to me. In the
>> first round you enabled SHAx_HWACCEL, which is broken on some architectures. That's the reason,
>> why is disabled by default. Ok, so first apply looks like this:
> let me rephrase my commit message.
>
> The SHAx_HWACCEL is broken on some architectures, right.
> However, when you take the defconfig, without the fragments, it will currently be enabled.
> Regardless of what's in my patch.
>
> Why? Because CONFIG_SHA1_HWACCEL in Busybox' kconfig is 'default y', has no dependency on other
> options and is not listed in the 'defconfig' file so far.
>
> Thus, my first patch should not change the current situation.
Ok, I see the point. So basically you can drop the sha1sum.cfg and 
sha256sum.cfg and merge it to defconfing. The result will be, that the 
acceleration is disabled.
>>    # CONFIG_FEATURE_BUFFERS_GO_IN_BSS is not set
>>    CONFIG_PASSWORD_MINLEN=6
>>    CONFIG_MD5_SMALL=1
>>   +CONFIG_SHA1_SMALL=3
>>   +CONFIG_SHA1_HWACCEL=y
>>   +CONFIG_SHA256_HWACCEL=y
>>    CONFIG_SHA3_SMALL=1
>>    CONFIG_FEATURE_NON_POSIX_CP=y
>>    # CONFIG_FEATURE_VERBOSE_CP_MESSAGE is not set
>>   @@ -884,6 +887,7 @@ CONFIG_FEATURE_HWIB=y
>>   
>>   Then I applied next patch:
>>   
>>    # CONFIG_FEATURE_BUFFERS_GO_IN_BSS is not set
>>    CONFIG_PASSWORD_MINLEN=6
>>    CONFIG_MD5_SMALL=1
>>   +CONFIG_SHA1_SMALL=3
>>   +CONFIG_SHA1_HWACCEL=y
>>   +CONFIG_SHA256_HWACCEL=y
>>    CONFIG_SHA3_SMALL=1
>>    CONFIG_FEATURE_NON_POSIX_CP=y
>>    # CONFIG_FEATURE_VERBOSE_CP_MESSAGE is not set
>>   @@ -278,8 +281,8 @@ CONFIG_FEATURE_LS_USERNAME=y
>>    CONFIG_FEATURE_LS_COLOR=y
>>    # CONFIG_FEATURE_LS_COLOR_IS_DEFAULT is not set
>>    CONFIG_MD5SUM=y
>>   -CONFIG_SHA1SUM=y
>>   -CONFIG_SHA256SUM=y
>>   +# CONFIG_SHA1SUM is not set
>>   +# CONFIG_SHA256SUM is not set
>>    # CONFIG_SHA512SUM is not set
>>    # CONFIG_SHA3SUM is not set
>>    
>>   @@ -383,8 +386,8 @@ CONFIG_DEFAULT_SETFONT_DIR=""
>>   
>>   and you disabled CONFIG_SHA1SUM, but keep enabled the acceleration? Is there any special reason
>> for that?
> There is one, because its the inverse of what the fragments apply, which is:
>
>    $ cat ./meta/recipes-core/busybox/busybox/sha1sum.cfg
>    CONFIG_SHA1SUM=y
>    CONFIG_SHA1_SMALL=3
>    # CONFIG_SHA1_HWACCEL is not set
>
>    $ cat ./meta/recipes-core/busybox/busybox/sha256sum.cfg
>    CONFIG_SHA256SUM=y
>    # CONFIG_SHA256_HWACCEL is not set
>
> I'm pretty sure not having touched the fragments.
>
> The CONFIG_SHA1_HWACCEL options could have been changed in the defconfig, too.
> But this is nothing that I've brought in. It's a result of
> https://github.com/openembedded/openembedded-core/commit/21753f16a364e32050cf8d79bfa7e0f89be52ce7
>
> Thus I'd assume my changes are consistent. Am I wrong?
No no, it's ok.
>
>>   Anyway, there was one commit in the usptream which was touching the acceleration. So I have to
>> try it on the broken machine and see if it's still broken or not.
> If you refer to oe-core upstream, yes there have been changes.
> This is why I needed to rebase and rework my branch.
> Or are there other changes I am not aware of, yet?
Not sure if you caught the point. I mean, that I will take a look on the 
upstream fixes for acceleration and then we can enable it by default.
>>> On Fri, 2024-10-25 at 19:26 +0200, Andrej wrote:
>>>   
>>>> Ok, I was thinking that it was done by intention, but it happened 14y
>>>> ago
>>>> https://github.com/openembedded/openembedded-core/commit/615a98ed9a02
>>>> 1da245513790c064761a0a5a67e9 . So I guess not.
>>   
>>> I am not sure if I fully understand what you are referring to here.
>>> Could you provide some details about what happened there?
>>> OE didn't have those fragments 14y ago I guess.
>>   So the idea behind is that you're going to remove the fragments and keep just a one defconfig.
>> But according to your commits you kept the fragments there, even if you put some values into
>> defconfig.
> Still not sure. I had no intention of removing the fragments.
> And I'm unsure how this relates to the referenced commit.
> My intention was to actually make them useful again, which is quite the opposite I guess.
>
>
> Regards, Enrico
>
>>   Regards,
>>   Andrej
>>
>>   
>> On 28.10.2024 09:29, Enrico Jörns wrote:
>>   
>>> This disables options from the defconfig file that are enabled by
>>> explicit config fragments.
>>>
>>> Having them enabled in the defconfig renders the fragments useless and
>>> takes away the ability to disable options with
>>>
>>>    SRC_URI:remove = "<fragment>.cfg".
>>>
>>> The respective options were all deactivated once but got accidentally
>>> enabled in 4335cd24 ("busybox: refresh the defconfig from 1.33.0").
>>>
>>> This commit disables the features for:
>>>
>>> - sha1sum.cfg
>>> - sha256sum.cfg
>>> - resize.cfg
>>> - pgrep.cfg
>>> - rev.cfg
>>>
>>> Signed-off-by: Enrico Jörns <ejo@pengutronix.de>
>>> ---
>>>   meta/recipes-core/busybox/busybox/defconfig | 12 ++++++------
>>>   1 file changed, 6 insertions(+), 6 deletions(-)
>>>
>>> diff --git a/meta/recipes-core/busybox/busybox/defconfig b/meta/recipes-
>>> core/busybox/busybox/defconfig
>>> index 6c5a8764ce..0ec65c65af 100644
>>> --- a/meta/recipes-core/busybox/busybox/defconfig
>>> +++ b/meta/recipes-core/busybox/busybox/defconfig
>>> @@ -281,8 +281,8 @@ CONFIG_FEATURE_LS_USERNAME=y
>>>   CONFIG_FEATURE_LS_COLOR=y
>>>   # CONFIG_FEATURE_LS_COLOR_IS_DEFAULT is not set
>>>   CONFIG_MD5SUM=y
>>> -CONFIG_SHA1SUM=y
>>> -CONFIG_SHA256SUM=y
>>> +# CONFIG_SHA1SUM is not set
>>> +# CONFIG_SHA256SUM is not set
>>>   # CONFIG_SHA512SUM is not set
>>>   # CONFIG_SHA3SUM is not set
>>>   
>>> @@ -386,8 +386,8 @@ CONFIG_DEFAULT_SETFONT_DIR=""
>>>   CONFIG_LOADKMAP=y
>>>   CONFIG_OPENVT=y
>>>   CONFIG_RESET=y
>>> -CONFIG_RESIZE=y
>>> -CONFIG_FEATURE_RESIZE_PRINT=y
>>> +# CONFIG_RESIZE is not set
>>> +# CONFIG_FEATURE_RESIZE_PRINT is not set
>>>   CONFIG_SETCONSOLE=y
>>>   # CONFIG_FEATURE_SETCONSOLE_LONG_OPTIONS is not set
>>>   # CONFIG_SETKEYCODES is not set
>>> @@ -689,7 +689,7 @@ CONFIG_RDATE=y
>>>   # CONFIG_RDEV is not set
>>>   # CONFIG_READPROFILE is not set
>>>   CONFIG_RENICE=y
>>> -CONFIG_REV=y
>>> +# CONFIG_REV is not set
>>>   # CONFIG_RTCWAKE is not set
>>>   # CONFIG_SCRIPT is not set
>>>   # CONFIG_SCRIPTREPLAY is not set
>>> @@ -1070,7 +1070,7 @@ CONFIG_KILLALL=y
>>>   # CONFIG_LSOF is not set
>>>   # CONFIG_MPSTAT is not set
>>>   # CONFIG_NMETER is not set
>>> -CONFIG_PGREP=y
>>> +# CONFIG_PGREP is not set
>>>   # CONFIG_PKILL is not set
>>>   CONFIG_PIDOF=y
>>>   # CONFIG_FEATURE_PIDOF_SINGLE is not set
>>   
>>   
So if the result is, that the configuration hasn't been changed at the 
end, I'm fine with the changes you did ;).

Regards,
Andrej

Re: [OE-core] [PATCH v2 2/9] busybox: disable defconfig options from simpler fragments

[Enrico Jörns]

Hi Andrej,

thanks for your review!

Am Dienstag, dem 29.10.2024 um 17:32 +0100 schrieb Andrej Valek:
> Hello Enrico,
> 
> On 28.10.2024 11:39, Enrico Jörns wrote:
> > Hej Andrej,
> > 
> > Am Montag, dem 28.10.2024 um 10:06 +0100 schrieb Andrej Valek:
> > >   Hello Enrico,
> > >   
> > >   I tried to follow (and applied) your patch, but it looks little bit inconsistent to me. In
> > > the
> > > first round you enabled SHAx_HWACCEL, which is broken on some architectures. That's the
> > > reason,
> > > why is disabled by default. Ok, so first apply looks like this:
> > let me rephrase my commit message.
> > 
> > The SHAx_HWACCEL is broken on some architectures, right.
> > However, when you take the defconfig, without the fragments, it will currently be enabled.
> > Regardless of what's in my patch.
> > 
> > Why? Because CONFIG_SHA1_HWACCEL in Busybox' kconfig is 'default y', has no dependency on other
> > options and is not listed in the 'defconfig' file so far.
> > 
> > Thus, my first patch should not change the current situation.
> Ok, I see the point. So basically you can drop the sha1sum.cfg and 
> sha256sum.cfg and merge it to defconfing. The result will be, that the 
> acceleration is disabled.

Just to get it right: Do we talk about the HWACCEL only or about enabling SHA1SUM and SHA256SUM
unconditionally and fully removing the fragments?

Personally, I'd be fine with this since the algorithm are quite essential anyway.

If that's the only thing to change, I'd start making a v3 for it.

> > 
> > >   Anyway, there was one commit in the usptream which was touching the acceleration. So I have
> > > to
> > > try it on the broken machine and see if it's still broken or not.
> > If you refer to oe-core upstream, yes there have been changes.
> > This is why I needed to rebase and rework my branch.
> > Or are there other changes I am not aware of, yet?
> Not sure if you caught the point. I mean, that I will take a look on the 
> upstream fixes for acceleration and then we can enable it by default.

Ok, so you referred to *busybox* upstream, not oe-core. Got it.

I've found https://git.busybox.net/busybox/commit/?id=bf57f732a5b6842f6fa3e0f90385f039e5d6a92c
Is this the fix you referred to?

> > > > [...]
> > >   
> So if the result is, that the configuration hasn't been changed at the 
> end, I'm fine with the changes you did ;).

Sounds good!

Regards, Enrico

> Regards,
> Andrej
> 

-- 
Pengutronix e.K.                           | Enrico Jörns                |
Embedded Linux Consulting & Support        | https://www.pengutronix.de/ |
Steuerwalder Str. 21                       | Phone: +49-5121-206917-180  |
31137 Hildesheim, Germany                  | Fax:   +49-5121-206917-9    |

RE: [OE-core] [PATCH v2 2/9] busybox: disable defconfig options from simpler fragments

[Peter Kjellerstedt]

> -----Original Message-----
> From: openembedded-core@lists.openembedded.org <openembedded-core@lists.openembedded.org> On Behalf Of Enrico Jörns
> Sent: den 29 oktober 2024 21:47
> To: Andrej Valek <andrej.v@skyrain.eu>; Richard Purdie <richard.purdie@linuxfoundation.org>
> Cc: yocto@pengutronix.de; openembedded-core@lists.openembedded.org
> Subject: Re: [OE-core] [PATCH v2 2/9] busybox: disable defconfig options from simpler fragments
> 
> Hi Andrej,
> 
> thanks for your review!
> 
> Am Dienstag, dem 29.10.2024 um 17:32 +0100 schrieb Andrej Valek:
> > Hello Enrico,
> >
> > On 28.10.2024 11:39, Enrico Jörns wrote:
> > > Hej Andrej,
> > >
> > > Am Montag, dem 28.10.2024 um 10:06 +0100 schrieb Andrej Valek:
> > > >   Hello Enrico,
> > > >
> > > >   I tried to follow (and applied) your patch, but it looks little
> bit inconsistent to me. In
> > > > the
> > > > first round you enabled SHAx_HWACCEL, which is broken on some
> architectures. That's the
> > > > reason,
> > > > why is disabled by default. Ok, so first apply looks like this:
> > > let me rephrase my commit message.
> > >
> > > The SHAx_HWACCEL is broken on some architectures, right.
> > > However, when you take the defconfig, without the fragments, it will
> currently be enabled.
> > > Regardless of what's in my patch.
> > >
> > > Why? Because CONFIG_SHA1_HWACCEL in Busybox' kconfig is 'default y',
> has no dependency on other
> > > options and is not listed in the 'defconfig' file so far.
> > >
> > > Thus, my first patch should not change the current situation.
> > Ok, I see the point. So basically you can drop the sha1sum.cfg and
> > sha256sum.cfg and merge it to defconfing. The result will be, that the
> > acceleration is disabled.
> 
> Just to get it right: Do we talk about the HWACCEL only or about enabling
> SHA1SUM and SHA256SUM
> unconditionally and fully removing the fragments?
> 
> Personally, I'd be fine with this since the algorithm are quite essential
> anyway.
> 
> If that's the only thing to change, I'd start making a v3 for it.
> 
> > >
> > > >   Anyway, there was one commit in the usptream which was touching
> the acceleration. So I have
> > > > to
> > > > try it on the broken machine and see if it's still broken or not.
> > > If you refer to oe-core upstream, yes there have been changes.
> > > This is why I needed to rebase and rework my branch.
> > > Or are there other changes I am not aware of, yet?
> > Not sure if you caught the point. I mean, that I will take a look on the
> > upstream fixes for acceleration and then we can enable it by default.
> 
> Ok, so you referred to *busybox* upstream, not oe-core. Got it.
> 
> I've found
> https://git.busybox.net/busybox/commit/?id=bf57f732a5b6842f6fa3e0f90385f039e5d6a92c
> Is this the fix you referred to?

I have sent a patch to backport the above commit. I also sent two patches to 
remove the no longer needed sha_accel.cfg file, and to clean up the sha*sum.cfg 
files.

> > > > > [...]
> > > >
> > So if the result is, that the configuration hasn't been changed at the
> > end, I'm fine with the changes you did ;).
> 
> Sounds good!
> 
> Regards, Enrico
> 
> > Regards,
> > Andrej

//Peter