Skip to contentSkip to conversation

Patch review

PATCH In review Steve Sakoman 3 participants 18 messages Aug 14, 2024 – Nov 5, 2024 1 message not in the archive

lore
Share & tools
lore
#0 lore raw

cover Patch review

Please review this set of changes for scarthgap and have comments back by end of day Friday, August 16

Passed a-full on autobuilder:

https://autobuilder.yoctoproject.org/typhoon/#/builders/83/builds/7241

The following changes since commit 236ac1b43308df722a78d3aa20aef065dfae5b2b:

build-appliance-image: Update to scarthgap head revision (2024-08-10 06:35:20 -0700)

are available in the Git repository at:

https://git.openembedded.org/openembedded-core-contrib stable/scarthgap-nut
https://git.openembedded.org/openembedded-core-contrib/log/?h=stable/scarthgap-nut

Archana Polampalli (1):
ffmpeg: fix CVE-2023-50008

Niko Mauno (3):
systemd: Mitigate /var/log type mismatch issue
systemd: Mitigate /var/tmp type mismatch issue image_types.bbclass: Use --force also with lz4,lzop

Peter Marko (1):
libyaml: ignore CVE-2024-35326

Richard Purdie (1):
cve_check: Use a local copy of the database during builds

Ross Burton (1):
python3-pycryptodome(x): use python_setuptools_build_meta build class

Ryan Eatmon (1):
u-boot.inc: Refactor do_* steps into functions that can be overridden

Soumya Sambu (1):
python3-certifi: Fix CVE-2024-39689

 meta/classes-recipe/image_types.bbclass       |   4 +-
 meta/classes/cve-check.bbclass                |   7 +-
 meta/recipes-bsp/u-boot/u-boot-configure.inc  |  36 ++-
 meta/recipes-bsp/u-boot/u-boot.inc            | 281 ++++++++++++------
 .../meta/cve-update-nvd2-native.bb            |  18 +-
 .../systemd/systemd/00-create-volatile.conf   |   1 +
 meta/recipes-core/systemd/systemd_255.4.bb    |   5 +-
 .../python3-certifi/CVE-2024-39689.patch      |  69 +++++
 .../python/python3-certifi_2024.2.2.bb        |   3 +
 .../python/python3-pycryptodome_3.20.0.bb     |   2 +-
 .../python/python3-pycryptodomex_3.20.0.bb    |   2 +-
 .../ffmpeg/ffmpeg/CVE-2023-50008.patch        |  29 ++
 .../recipes-multimedia/ffmpeg/ffmpeg_6.1.1.bb |   1 +
 meta/recipes-support/libyaml/libyaml_0.2.5.bb |   1 +
 14 files changed, 336 insertions(+), 123 deletions(-)
 create mode 100644 meta/recipes-devtools/python/python3-certifi/CVE-2024-39689.patch
 create mode 100644 meta/recipes-multimedia/ffmpeg/ffmpeg/CVE-2023-50008.patch
Signature · 1 line
2.34.1
#1 lore raw

PATCH 1/9 cve_check: Use a local copy of the database during builds

From: Richard Purdie <[email protected]>

Rtaher than trying to use a sqlite database over NFS from DL_DIR, work from a local copy in STAGING DIR after fetching.

(cherry picked from commit 03596904392d257572a905a182b92c780d636744)
---
 meta/classes/cve-check.bbclass                 |  7 ++++---
 .../meta/cve-update-nvd2-native.bb             | 18 +++++++++++++-----
 2 files changed, 17 insertions(+), 8 deletions(-)
meta/classes/cve-check.bbclass +4 -3
index 93a2a1413d..c946de29a4 100644
@@ -31,8 +31,9 @@
 CVE_PRODUCT ??= "${BPN}"
 CVE_VERSION ??= "${PV}"
 
-CVE_CHECK_DB_DIR ?= "${DL_DIR}/CVE_CHECK"
-CVE_CHECK_DB_FILE ?= "${CVE_CHECK_DB_DIR}/nvdcve_2-1.db"
+CVE_CHECK_DB_FILENAME ?= "nvdcve_2-1.db"
+CVE_CHECK_DB_DIR ?= "${STAGING_DIR}/CVE_CHECK"
+CVE_CHECK_DB_FILE ?= "${CVE_CHECK_DB_DIR}/${CVE_CHECK_DB_FILENAME}"
 CVE_CHECK_DB_FILE_LOCK ?= "${CVE_CHECK_DB_FILE}.lock"
 
 CVE_CHECK_LOG ?= "${T}/cve.log"
@@ -198,7 +199,7 @@ python do_cve_check () {
 }
 
 addtask cve_check before do_build
-do_cve_check[depends] = "cve-update-nvd2-native:do_fetch"
+do_cve_check[depends] = "cve-update-nvd2-native:do_unpack"
 do_cve_check[nostamp] = "1"
 
 python cve_check_cleanup () {
meta/recipes-core/meta/cve-update-nvd2-native.bb +13 -5
index 1901641965..2d23d28c3e 100644
@@ -8,7 +8,6 @@ INHIBIT_DEFAULT_DEPS = "1"
 
 inherit native
 
-deltask do_unpack
 deltask do_patch
 deltask do_configure
 deltask do_compile
@@ -35,7 +34,9 @@ CVE_DB_INCR_UPDATE_AGE_THRES ?= "10368000"
 # Number of attempts for each http query to nvd server before giving up
 CVE_DB_UPDATE_ATTEMPTS ?= "5"
 
-CVE_DB_TEMP_FILE ?= "${CVE_CHECK_DB_DIR}/temp_nvdcve_2.db"
+CVE_CHECK_DB_DLDIR_FILE ?= "${DL_DIR}/CVE_CHECK/${CVE_CHECK_DB_FILENAME}"
+CVE_CHECK_DB_DLDIR_LOCK ?= "${CVE_CHECK_DB_DLDIR_FILE}.lock"
+CVE_CHECK_DB_TEMP_FILE ?= "${CVE_CHECK_DB_FILE}.tmp"
 
 python () {
     if not bb.data.inherits_class("cve-check", d):
@@ -52,9 +53,9 @@ python do_fetch() {
 
     bb.utils.export_proxies(d)
 
-    db_file = d.getVar("CVE_CHECK_DB_FILE")
+    db_file = d.getVar("CVE_CHECK_DB_DLDIR_FILE")
     db_dir = os.path.dirname(db_file)
-    db_tmp_file = d.getVar("CVE_DB_TEMP_FILE")
+    db_tmp_file = d.getVar("CVE_CHECK_DB_TEMP_FILE")
 
     cleanup_db_download(db_file, db_tmp_file)
     # By default let's update the whole database (since time 0)
@@ -77,6 +78,7 @@ python do_fetch() {
         pass
 
     bb.utils.mkdirhier(db_dir)
+    bb.utils.mkdirhier(os.path.dirname(db_tmp_file))
     if os.path.exists(db_file):
         shutil.copy2(db_file, db_tmp_file)
 
@@ -89,10 +91,16 @@ python do_fetch() {
         os.remove(db_tmp_file)
 }
 
-do_fetch[lockfiles] += "${CVE_CHECK_DB_FILE_LOCK}"
+do_fetch[lockfiles] += "${CVE_CHECK_DB_DLDIR_LOCK}"
 do_fetch[file-checksums] = ""
 do_fetch[vardeps] = ""
 
+python do_unpack() {
+    import shutil
+    shutil.copyfile(d.getVar("CVE_CHECK_DB_DLDIR_FILE"), d.getVar("CVE_CHECK_DB_FILE"))
+}
+do_unpack[lockfiles] += "${CVE_CHECK_DB_DLDIR_LOCK} ${CVE_CHECK_DB_FILE_LOCK}"
+
 def cleanup_db_download(db_file, db_tmp_file):
     """
     Cleanup the download space from possible failed downloads
Signature · 1 line
2.34.1
#11 lore raw

PATCH 2/9 libyaml: ignore CVE-2024-35326

In reply to Steve Sakoman

From: Peter Marko <[email protected]>

This is the same problem as already ignored CVE-2024-35328. See laso this comment in addition:
https://github.com/yaml/libyaml/issues/298#issuecomment-2167684233

---
 meta/recipes-support/libyaml/libyaml_0.2.5.bb | 1 +
 1 file changed, 1 insertion(+)
meta/recipes-support/libyaml/libyaml_0.2.5.bb +1 -0
index 1c6a5fcb45..334d9113d2 100644
@@ -18,6 +18,7 @@ inherit autotools
 DISABLE_STATIC:class-nativesdk = ""
 DISABLE_STATIC:class-native = ""
 
+CVE_STATUS[CVE-2024-35326] = "upstream-wontfix: Upstream thinks there is no working code that is exploitable - https://github.com/yaml/libyaml/issues/302"
 CVE_STATUS[CVE-2024-35328] = "upstream-wontfix: Upstream thinks there is no working code that is exploitable - https://github.com/yaml/libyaml/issues/302"
 
 BBCLASSEXTEND = "native nativesdk"
Signature · 1 line
2.34.1
#12 lore raw

PATCH 3/9 python3-certifi: Fix CVE-2024-39689

In reply to Steve Sakoman

From: Soumya Sambu <[email protected]>

Certifi is a curated collection of Root Certificates for validating the trustworthiness of SSL certificates while verifying the identity of TLS hosts. Certifi starting in 2021.05.30 and prior to 2024.07.4 recognized root certificates from GLOBALTRUST. Certifi 2024.07.04 removes root certificates from GLOBALTRUST from the root store. These are in the process of being removed from Mozilla's trust store. GLOBALTRUST's root certificates are being removed pursuant to an investigation which identified "long-running and unresolved compliance issues."Certifi is a curated collection of Root Certificates for validating the trustworthiness of SSL certificates while verifying the identity of TLS hosts. Certifi starting in 2021.05.30 and prior to 2024.07.4 recognized root certificates from GLOBALTRUST. Certifi 2024.07.04 removes root certificates from GLOBALTRUST from the root store. These are in the process of being removed from Mozilla's trust store. GLOBALTRUST's root certificates are being removed pursuant to an investigation which identified "long-running and unresolved compliance issues."

References:
https://nvd.nist.gov/vuln/detail/CVE-2024-39689

Upstream-patch:
https://github.com/certifi/python-certifi/commit/bd8153872e9c6fc98f4023df9c2deaffea2fa463

---
 .../python3-certifi/CVE-2024-39689.patch      | 69 +++++++++++++++++++
 .../python/python3-certifi_2024.2.2.bb        |  3 +
 2 files changed, 72 insertions(+)
 create mode 100644 meta/recipes-devtools/python/python3-certifi/CVE-2024-39689.patch
meta/recipes-devtools/python/python3-certifi/CVE-2024-39689.patch new file +69 -0
new file mode 100644
index 0000000000..a2ecc15d2c
@@ -0,0 +1,69 @@
+From bd8153872e9c6fc98f4023df9c2deaffea2fa463 Mon Sep 17 00:00:00 2001
+From: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
+Date: Wed, 3 Jul 2024 21:34:29 -0400
+Subject: [PATCH] 2024.07.04 (#295)
+
+Co-authored-by: alex <[email protected]>
+
+CVE: CVE-2024-39689
+
+Upstream-Status: Backport [https://github.com/certifi/python-certifi/commit/bd8153872e9c6fc98f4023df9c2deaffea2fa463]
+
+Signed-off-by: Soumya Sambu <[email protected]>
+---
+ certifi/cacert.pem | 40 ----------------------------------------
+ 1 file changed, 40 deletions(-)
+
+diff --git a/certifi/cacert.pem b/certifi/cacert.pem
+index 1bec256..6bb8cf8 100644
+--- a/certifi/cacert.pem
++++ b/certifi/cacert.pem
+@@ -3857,46 +3857,6 @@ DgQWBBQxCpCPtsad0kRLgLWi5h+xEk8blTAKBggqhkjOPQQDAwNoADBlAjEA31SQ
+ +RHUjE7AwWHCFUyqqx0LMV87HOIAl0Qx5v5zli/altP+CAezNIm8BZ/3Hobui3A=
+ -----END CERTIFICATE-----
+
+-# Issuer: CN=GLOBALTRUST 2020 O=e-commerce monitoring GmbH
+-# Subject: CN=GLOBALTRUST 2020 O=e-commerce monitoring GmbH
+-# Label: "GLOBALTRUST 2020"
+-# Serial: 109160994242082918454945253
+-# MD5 Fingerprint: 8a:c7:6f:cb:6d:e3:cc:a2:f1:7c:83:fa:0e:78:d7:e8
+-# SHA1 Fingerprint: d0:67:c1:13:51:01:0c:aa:d0:c7:6a:65:37:31:16:26:4f:53:71:a2
+-# SHA256 Fingerprint: 9a:29:6a:51:82:d1:d4:51:a2:e3:7f:43:9b:74:da:af:a2:67:52:33:29:f9:0f:9a:0d:20:07:c3:34:e2:3c:9a
+------BEGIN CERTIFICATE-----
+-MIIFgjCCA2qgAwIBAgILWku9WvtPilv6ZeUwDQYJKoZIhvcNAQELBQAwTTELMAkG
+-A1UEBhMCQVQxIzAhBgNVBAoTGmUtY29tbWVyY2UgbW9uaXRvcmluZyBHbWJIMRkw
+-FwYDVQQDExBHTE9CQUxUUlVTVCAyMDIwMB4XDTIwMDIxMDAwMDAwMFoXDTQwMDYx
+-MDAwMDAwMFowTTELMAkGA1UEBhMCQVQxIzAhBgNVBAoTGmUtY29tbWVyY2UgbW9u
+-aXRvcmluZyBHbWJIMRkwFwYDVQQDExBHTE9CQUxUUlVTVCAyMDIwMIICIjANBgkq
+-hkiG9w0BAQEFAAOCAg8AMIICCgKCAgEAri5WrRsc7/aVj6B3GyvTY4+ETUWiD59b
+-RatZe1E0+eyLinjF3WuvvcTfk0Uev5E4C64OFudBc/jbu9G4UeDLgztzOG53ig9Z
+-YybNpyrOVPu44sB8R85gfD+yc/LAGbaKkoc1DZAoouQVBGM+uq/ufF7MpotQsjj3
+-QWPKzv9pj2gOlTblzLmMCcpL3TGQlsjMH/1WljTbjhzqLL6FLmPdqqmV0/0plRPw
+-yJiT2S0WR5ARg6I6IqIoV6Lr/sCMKKCmfecqQjuCgGOlYx8ZzHyyZqjC0203b+J+
+-BlHZRYQfEs4kUmSFC0iAToexIiIwquuuvuAC4EDosEKAA1GqtH6qRNdDYfOiaxaJ
+-SaSjpCuKAsR49GiKweR6NrFvG5Ybd0mN1MkGco/PU+PcF4UgStyYJ9ORJitHHmkH
+-r96i5OTUawuzXnzUJIBHKWk7buis/UDr2O1xcSvy6Fgd60GXIsUf1DnQJ4+H4xj0
+-4KlGDfV0OoIu0G4skaMxXDtG6nsEEFZegB31pWXogvziB4xiRfUg3kZwhqG8k9Me
+-dKZssCz3AwyIDMvUclOGvGBG85hqwvG/Q/lwIHfKN0F5VVJjjVsSn8VoxIidrPIw
+-q7ejMZdnrY8XD2zHc+0klGvIg5rQmjdJBKuxFshsSUktq6HQjJLyQUp5ISXbY9e2
+-nKd+Qmn7OmMCAwEAAaNjMGEwDwYDVR0TAQH/BAUwAwEB/zAOBgNVHQ8BAf8EBAMC
+-AQYwHQYDVR0OBBYEFNwuH9FhN3nkq9XVsxJxaD1qaJwiMB8GA1UdIwQYMBaAFNwu
+-H9FhN3nkq9XVsxJxaD1qaJwiMA0GCSqGSIb3DQEBCwUAA4ICAQCR8EICaEDuw2jA
+-VC/f7GLDw56KoDEoqoOOpFaWEhCGVrqXctJUMHytGdUdaG/7FELYjQ7ztdGl4wJC
+-XtzoRlgHNQIw4Lx0SsFDKv/bGtCwr2zD/cuz9X9tAy5ZVp0tLTWMstZDFyySCstd
+-6IwPS3BD0IL/qMy/pJTAvoe9iuOTe8aPmxadJ2W8esVCgmxcB9CpwYhgROmYhRZf
+-+I/KARDOJcP5YBugxZfD0yyIMaK9MOzQ0MAS8cE54+X1+NZK3TTN+2/BT+MAi1bi
+-kvcoskJ3ciNnxz8RFbLEAwW+uxF7Cr+obuf/WEPPm2eggAe2HcqtbepBEX4tdJP7
+-wry+UUTF72glJ4DjyKDUEuzZpTcdN3y0kcra1LGWge9oXHYQSa9+pTeAsRxSvTOB
+-TI/53WXZFM2KJVj04sWDpQmQ1GwUY7VA3+vA/MRYfg0UFodUJ25W5HCEuGwyEn6C
+-MUO+1918oa2u1qsgEu8KwxCMSZY13At1XrFP1U80DhEgB3VDRemjEdqso5nCtnkn
+-4rnvyOL2NSl6dPrFf4IFYqYK6miyeUcGbvJXqBUzxvd4Sj1Ce2t+/vdG6tHrju+I
+-aFvowdlxfv1k7/9nR4hYJS8+hge9+6jlgqispdNpQ80xiEmEU5LAsTkbOYMBMMTy
+-qfrQA71yN2BWHzZ8vTmR9W0Nv3vXkg==
+------END CERTIFICATE-----
+-
+ # Issuer: CN=ANF Secure Server Root CA O=ANF Autoridad de Certificacion OU=ANF CA Raiz
+ # Subject: CN=ANF Secure Server Root CA O=ANF Autoridad de Certificacion OU=ANF CA Raiz
+ # Label: "ANF Secure Server Root CA"
+--
+2.40.0
meta/recipes-devtools/python/python3-certifi_2024.2.2.bb +3 -0
index 4e61b8d9d4..116add2079 100644
@@ -7,6 +7,9 @@ HOMEPAGE = " http://certifi.io/"
 LICENSE = "ISC"
 LIC_FILES_CHKSUM = "file://LICENSE;md5=11618cb6a975948679286b1211bd573c"
 
+SRC_URI += "file://CVE-2024-39689.patch \
+           "
+
 SRC_URI[sha256sum] = "0569859f95fc761b18b45ef421b1290a0f65f147e92a1e5eb3e635f9a5e4e66f"
 
 inherit pypi setuptools3
Signature · 1 line
2.34.1
#13 lore raw

PATCH 4/9 ffmpeg: fix CVE-2023-50008

In reply to Steve Sakoman

From: Archana Polampalli <[email protected]>

Buffer Overflow vulnerability in Ffmpeg v.n6.1-3-g466799d4f5 allows a local attacker to execute arbitrary code via the av_malloc function in libavutil/mem.c:105:9 component.

---
 .../ffmpeg/ffmpeg/CVE-2023-50008.patch        | 29 +++++++++++++++++++
 .../recipes-multimedia/ffmpeg/ffmpeg_6.1.1.bb |  1 +
 2 files changed, 30 insertions(+)
 create mode 100644 meta/recipes-multimedia/ffmpeg/ffmpeg/CVE-2023-50008.patch
meta/recipes-multimedia/ffmpeg/ffmpeg/CVE-2023-50008.patch new file +29 -0
new file mode 100644
index 0000000000..4b8935628f
@@ -0,0 +1,29 @@
+From 5f87a68cf70dafeab2fb89b42e41a4c29053b89b Mon Sep 17 00:00:00 2001
+From: Paul B Mahol <[email protected]>
+Date: Mon, 27 Nov 2023 12:08:20 +0100
+Subject: [PATCH] avfilter/vf_colorcorrect: fix memory leaks
+
+CVE: CVE-2023-50008
+
+Upstream-Status: Backport [https://github.com/FFmpeg/FFmpeg/commit/5f87a68cf70dafeab2fb89b42e41a4c29053b89b]
+
+Signed-off-by: Archana Polampalli <[email protected]>
+---
+ libavfilter/vf_colorcorrect.c | 2 ++
+ 1 file changed, 2 insertions(+)
+
+diff --git a/libavfilter/vf_colorcorrect.c b/libavfilter/vf_colorcorrect.c
+index 1c4dea5..6bdec2c 100644
+--- a/libavfilter/vf_colorcorrect.c
++++ b/libavfilter/vf_colorcorrect.c
+@@ -497,6 +497,8 @@ static av_cold void uninit(AVFilterContext *ctx)
+     ColorCorrectContext *s = ctx->priv;
+
+     av_freep(&s->analyzeret);
++    av_freep(&s->uhistogram);
++    av_freep(&s->vhistogram);
+ }
+
+ static const AVFilterPad colorcorrect_inputs[] = {
+--
+2.40.0
meta/recipes-multimedia/ffmpeg/ffmpeg_6.1.1.bb +1 -0
index 05919e6ffd..9b74d78fb1 100644
@@ -30,6 +30,7 @@ SRC_URI = " \
     file://CVE-2023-49502.patch \
     file://CVE-2024-31578.patch \
     file://CVE-2024-31582.patch \
+    file://CVE-2023-50008.patch \
 "
 
 SRC_URI[sha256sum] = "8684f4b00f94b85461884c3719382f1261f0d9eb3d59640a1f4ac0873616f968"
Signature · 1 line
2.34.1
#14 lore raw

PATCH 5/9 python3-pycryptodome(x): use python_setuptools_build_meta build class

In reply to Steve Sakoman

From: Ross Burton <[email protected]>

This package can be built using pep517 classes now.

(cherry picked from commit a32fa3e64d1daf5846c29403e9f258aea42212d3)
---
 meta/recipes-devtools/python/python3-pycryptodome_3.20.0.bb  | 2 +-
 meta/recipes-devtools/python/python3-pycryptodomex_3.20.0.bb | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)
meta/recipes-devtools/python/python3-pycryptodome_3.20.0.bb +1 -1
index d24fa58d43..6c93c205ac 100644
@@ -1,5 +1,5 @@
 require python-pycryptodome.inc
-inherit setuptools3
+inherit python_setuptools_build_meta
 
 SRC_URI[sha256sum] = "09609209ed7de61c2b560cc5c8c4fbf892f8b15b1faf7e4cbffac97db1fffda7"
 
meta/recipes-devtools/python/python3-pycryptodomex_3.20.0.bb +1 -1
index 2673ea8326..54578d2850 100644
@@ -1,5 +1,5 @@
 require python-pycryptodome.inc
-inherit setuptools3
+inherit python_setuptools_build_meta
 
 SRC_URI[sha256sum] = "7a710b79baddd65b806402e14766c721aee8fb83381769c27920f26476276c1e"
 
Signature · 1 line
2.34.1
#15 lore raw

PATCH 6/9 systemd: Mitigate /var/log type mismatch issue

In reply to Steve Sakoman

From: Niko Mauno <[email protected]>

When VOLATILE_LOG_DIR evaluates as True, the base-files recipe provides /var/log -> /var/volatile/log symlink which is in conflict with systemd upstream tmpfiles.d/var.conf.in which defines it as a directory.

This generates following error in journal:

  Jul 03 14:28:00 qemux86-64 systemd-tmpfiles[165]: "/var/log" already exists and is not a directory.

Mitigate the issue by defining /var/log as symlink corresponding to the one created by base-files, when appropriate.

(From OE-Core rev: 711ee36e88c8968e3c45ea787b3adcf64352adf9)

---
 meta/recipes-core/systemd/systemd_255.4.bb | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
meta/recipes-core/systemd/systemd_255.4.bb +3 -2
index f58a1bc2b6..0ccca8a567 100644
@@ -307,9 +307,10 @@ do_install() {
 	fi
 
 	if "${@'true' if oe.types.boolean(d.getVar('VOLATILE_LOG_DIR')) else 'false'}"; then
-		# /var/log is typically a symbolic link to inside /var/volatile,
-		# which is expected to be empty.
+		# base-files recipe provides /var/log which is a symlink to /var/volatile/log
 		rm -rf ${D}${localstatedir}/log
+		printf 'L\t\t%s/log\t\t-\t-\t-\t-\t%s/volatile/log\n' "${localstatedir}" \
+			"${localstatedir}" >>${D}${nonarch_libdir}/tmpfiles.d/00-create-volatile.conf
 	elif [ -e ${D}${localstatedir}/log/journal ]; then
 		chown root:systemd-journal ${D}${localstatedir}/log/journal
 
Signature · 1 line
2.34.1
#16 lore raw

PATCH 7/9 systemd: Mitigate /var/tmp type mismatch issue

In reply to Steve Sakoman

From: Niko Mauno <[email protected]>

The base-files recipe provides /var/tmp -> /var/volatile/tmp symlink which is in conflict with systemd upstream tmpfiles.d/tmp.conf which defines it as a directory (or subvolume on btrfs).

This generates following error in journal:

  Jul 03 15:37:21 qemux86-64 systemd-tmpfiles[158]: "/var/tmp" already exists and is not a directory.

Mitigate the issue by defining /var/tmp as symlink corresponding to the one created by base-files.

(From OE-Core rev: 1f1f6f45e3cfe24dfee8a09d01a5d32f3080e381)

---
 meta/recipes-core/systemd/systemd/00-create-volatile.conf | 1 +
 1 file changed, 1 insertion(+)
meta/recipes-core/systemd/systemd/00-create-volatile.conf +1 -0
index c4277221a2..043b2ef1d8 100644
@@ -6,3 +6,4 @@
 d		/run/lock		1777	-	-	-
 d		/var/volatile/log		-	-	-	-
 d		/var/volatile/tmp		1777	-	-
+L		/var/tmp		-	-	-	-	/var/volatile/tmp
Signature · 1 line
2.34.1
#17 lore raw

PATCH 8/9 image_types.bbclass: Use --force also with lz4,lzop

In reply to Steve Sakoman

From: Niko Mauno <[email protected]>

Several conversion commands already make use of 'force' option in the compression, which enables overwriting existing files without prompting.

Since occasionally an existing residual destination file from a previously aborted or failed task can prevent the re-execution of the conversion command task, by enabling the 'force' option also for lz4 and lzop compression commands we can avoid following kind of BitBake failures with these compressors:

  | DEBUG: Executing shell function do_image_cpio
  | 117685 blocks
  | 2 blocks
  | example-image.cpio.lz4 already exists; do you want to overwrite (y/N) ?     not overwritten
  | Error 20 : example-image.cpio : open file error
  | WARNING: exit code 20 from a shell command.

ERROR: Task (.../recipes-core/images/example-image.bb:do_image_cpio) failed with exit code '1'

(From OE-Core rev: 623ab22434909f10aaf613cd3032cc2a2c6e3ff9)

---
 meta/classes-recipe/image_types.bbclass | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
meta/classes-recipe/image_types.bbclass +2 -2
index 2f948ecbf8..28afff4571 100644
@@ -335,8 +335,8 @@ CONVERSION_CMD:lzma = "lzma -k -f -7 ${IMAGE_NAME}.${type}"
 CONVERSION_CMD:gz = "gzip -f -9 -n -c --rsyncable ${IMAGE_NAME}.${type} > ${IMAGE_NAME}.${type}.gz"
 CONVERSION_CMD:bz2 = "pbzip2 -f -k ${IMAGE_NAME}.${type}"
 CONVERSION_CMD:xz = "xz -f -k -c ${XZ_COMPRESSION_LEVEL} ${XZ_DEFAULTS} --check=${XZ_INTEGRITY_CHECK} ${IMAGE_NAME}.${type} > ${IMAGE_NAME}.${type}.xz"
-CONVERSION_CMD:lz4 = "lz4 -9 -z -l ${IMAGE_NAME}.${type} ${IMAGE_NAME}.${type}.lz4"
-CONVERSION_CMD:lzo = "lzop -9 ${IMAGE_NAME}.${type}"
+CONVERSION_CMD:lz4 = "lz4 -f -9 -z -l ${IMAGE_NAME}.${type} ${IMAGE_NAME}.${type}.lz4"
+CONVERSION_CMD:lzo = "lzop -f -9 ${IMAGE_NAME}.${type}"
 CONVERSION_CMD:zip = "zip ${ZIP_COMPRESSION_LEVEL} ${IMAGE_NAME}.${type}.zip ${IMAGE_NAME}.${type}"
 CONVERSION_CMD:7zip = "7za a -mx=${7ZIP_COMPRESSION_LEVEL} -mm=${7ZIP_COMPRESSION_METHOD} ${IMAGE_NAME}.${type}.${7ZIP_EXTENSION} ${IMAGE_NAME}.${type}"
 CONVERSION_CMD:zst = "zstd -f -k -c ${ZSTD_DEFAULTS} ${IMAGE_NAME}.${type} > ${IMAGE_NAME}.${type}.zst"
Signature · 1 line
2.34.1
#18 lore raw

PATCH 9/9 u-boot.inc: Refactor do_* steps into functions that can be overridden

In reply to Steve Sakoman

From: Ryan Eatmon <[email protected]>

The looping logic for handling (and not handling) UBOOT_CONFIG has led to the various do_* functions to be large and unwieldy. In order to modify one of the functional blocks inside of a loop (or in the else condition) means you either have to replace the function entirely, or append the function and undo something it did and then do what you need for your change.

This refactor breaks out all of the inner loops and else clauses into new functions that themselves can be overridden without needing to worry about the bulk of the looping logic.

It should not break any existing recipes doing prepends, appends, or overrides. None of the functional blocks were changed, just refactored out into new functions.

Backport from master: https://git.openembedded.org/openembedded-core/commit/?id=937bcc229502fcc154cc676b4fcc93c561873def

---
 meta/recipes-bsp/u-boot/u-boot-configure.inc |  36 ++-
 meta/recipes-bsp/u-boot/u-boot.inc           | 281 ++++++++++++-------
 2 files changed, 208 insertions(+), 109 deletions(-)
meta/recipes-bsp/u-boot/u-boot-configure.inc +24 -12
index 378d675364..a15511f8b2 100644
@@ -18,23 +18,35 @@ do_configure () {
             for type in ${UBOOT_CONFIG}; do
                 j=$(expr $j + 1);
                 if [ $j -eq $i ]; then
-                    oe_runmake -C ${S} O=${B}/${config} ${config}
-                    if [ -n "${@' '.join(find_cfgs(d))}" ]; then
-                        merge_config.sh -m -O ${B}/${config} ${B}/${config}/.config ${@" ".join(find_cfgs(d))}
-                        oe_runmake -C ${S} O=${B}/${config} oldconfig
-                    fi
+                    uboot_configure_config $config $type
                 fi
             done
             unset j
         done
         unset i
     else
-        if [ -n "${UBOOT_MACHINE}" ]; then
-            oe_runmake -C ${S} O=${B} ${UBOOT_MACHINE}
-        else
-            oe_runmake -C ${S} O=${B} oldconfig
-        fi
-        merge_config.sh -m .config ${@" ".join(find_cfgs(d))}
-        cml1_do_configure
+        uboot_configure
     fi
 }
+
+uboot_configure_config () {
+    config=$1
+    type=$2
+
+    oe_runmake -C ${S} O=${B}/${config} ${config}
+    if [ -n "${@' '.join(find_cfgs(d))}" ]; then
+        merge_config.sh -m -O ${B}/${config} ${B}/${config}/.config ${@" ".join(find_cfgs(d))}
+        oe_runmake -C ${S} O=${B}/${config} oldconfig
+    fi
+}
+
+uboot_configure () {
+    if [ -n "${UBOOT_MACHINE}" ]; then
+        oe_runmake -C ${S} O=${B} ${UBOOT_MACHINE}
+    else
+        oe_runmake -C ${S} O=${B} oldconfig
+    fi
+    merge_config.sh -m .config ${@" ".join(find_cfgs(d))}
+    cml1_do_configure
+}
+
meta/recipes-bsp/u-boot/u-boot.inc +184 -97

387 lines not on this page · raw message

Signature · 1 line
2.34.1
#2 lore raw

Re: [OE-core][scarthgap 1/9] cve_check: Use a local copy of the database during builds

In reply to Steve Sakoman
Quoted 9 lines · Steve Sakoman via lists.openembedded.org

From: Richard Purdie <[email protected]>

Rtaher than trying to use a sqlite database over NFS from DL_DIR, work from a local copy in STAGING DIR after fetching.

(cherry picked from commit 03596904392d257572a905a182b92c780d636744)

This patch is changing the ABI (well... variables). I think we should come back to the old naming
for the downloaded copy before backporting.

Regards,
Marta

#3 lore raw

Re: [OE-core][scarthgap 1/9] cve_check: Use a local copy of the database during builds

Thanks for reviewing Marta!

I'll drop this backported patch for both scarthgap and kirkstone and wait for appropriate versions of the patch to be submitted.

Steve

Quoted 21 lines · Marta Rybczynska
Quoted 9 lines · Steve Sakoman via lists.openembedded.org

From: Richard Purdie <[email protected]>

Rtaher than trying to use a sqlite database over NFS from DL_DIR, work from a local copy in STAGING DIR after fetching.

(cherry picked from commit 03596904392d257572a905a182b92c780d636744)

This patch is changing the ABI (well... variables). I think we should come back to the old naming for the downloaded copy before backporting.

Regards,
Marta

#5 lore raw

Re: [OE-core][scarthgap 1/9] cve_check: Use a local copy of the database during builds

In reply to a message not in the archive

Hi Marta,

We are still getting database errors on scarthgap (and occasionally on kirkstone if memory serves), so I think we need to reconsider backporting this patch.

Can you suggest changes that would make it acceptable to you?

Steve

Quoted 39 lines · Steve Sakoman via lists.openembedded.org

Thanks for reviewing Marta!

I'll drop this backported patch for both scarthgap and kirkstone and wait for appropriate versions of the patch to be submitted.

Steve

Quoted 16 lines · Marta Rybczynska
Quoted 9 lines · Steve Sakoman via lists.openembedded.org

From: Richard Purdie <[email protected]>

Rtaher than trying to use a sqlite database over NFS from DL_DIR, work from a local copy in STAGING DIR after fetching.

(cherry picked from commit 03596904392d257572a905a182b92c780d636744)
Quoted 5 lines

This patch is changing the ABI (well... variables). I think we should come back to the old naming for the downloaded copy before backporting.

Regards,
Marta

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#203323): https://lists.openembedded.org/g/openembedded-core/message/203323
Mute This Topic: https://lists.openembedded.org/mt/107893246/3620601
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-

#6 lore raw

Re: [OE-core][scarthgap 1/9] cve_check: Use a local copy of the database during builds

Quoted 7 lines · Steve Sakoman via lists.openembedded.org

Hi Marta,

We are still getting database errors on scarthgap (and occasionally on kirkstone if memory serves), so I think we need to reconsider backporting this patch.

Can you suggest changes that would make it acceptable to you?

Just for complete context, moving the database on master did isolate the database issues to scarthgap so the problem is in the older branches at this point.

Cheers,

Richard

#7 lore raw

Re: [OE-core][scarthgap 1/9] cve_check: Use a local copy of the database during builds

Quoted 8 lines · Richard Purdie
Quoted 6 lines · Steve Sakoman via lists.openembedded.org

Hi Marta,

We are still getting database errors on scarthgap (and occasionally on kirkstone if memory serves), so I think we need to reconsider backporting this patch.

Quoted 7 lines
Quoted 1 line

Can you suggest changes that would make it acceptable to you?

Just for complete context, moving the database on master did isolate the database issues to scarthgap so the problem is in the older branches at this point.

Hello all,
To make a summary, this patch did not solve the issue on master, it was the change of the database path that did. Is this correct? And we still do not know what the cause is? So, when we change the path on all branches, the issue is likely to come back ?

My problem with this patch is that it changes names of variables that are in use. It should at least not fail silently for people who decide to use a pinned database.

Do you see another solution than renaming the _DLDIR_ version to the old name?

Kind regards,
Marta

#8 lore raw

Re: [OE-core][scarthgap 1/9] cve_check: Use a local copy of the database during builds

Quoted 2 lines · Marta Rybczynska

To make a summary, this patch did not solve the issue on master, it was the change of the database path that did.

Originally, I couldn't isolate the change to tell or not. Now, by changing the database path we've shown the issue is occurring on the branches without the patch.

Quoted 3 lines

Is this correct? And we still do not know what the cause is? So,
when we change the path on all branches, the issue is likely to come back ?

I think if we move the database operations to the local filesystem and out of DL_DIR, we're likely to stop seeing the issues. We know they're isolated to something about the older releases.

Quoted 6 lines

My problem with this patch is that it changes names of variables that are in use. It should at least not fail silently for people who decide to use a pinned database.

Do you see another solution than renaming the _DLDIR_ version to the old name?

We either backport the patch as is and have the same functionality across the branches, or we do some special version of the patch for the LTS branches. In many ways I think it might be better to have things consistent rather than doing something special with all the confusion that can cause, even if people do have to actively adapt to the change.

Cheers,

Richard

#9 lore raw

Re: [OE-core][scarthgap 1/9] cve_check: Use a local copy of the database during builds

Quoted 20 lines · On Wed, 30 Oct 2024, 07:40 Richard Purdie, <…
Quoted 2 lines · Marta Rybczynska

To make a summary, this patch did not solve the issue on master, it was the change of the database path that did.

Originally, I couldn't isolate the change to tell or not. Now, by changing the database path we've shown the issue is occurring on the branches without the patch.

Quoted 3 lines

Is this correct? And we still do not know what the cause is? So,
when we change the path on all branches, the issue is likely to come back ?

I think if we move the database operations to the local filesystem and out of DL_DIR, we're likely to stop seeing the issues. We know they're isolated to something about the older releases.

Quoted 4 lines

My problem with this patch is that it changes names of variables that are in use. It should at least not fail silently for people who decide to use a pinned database.

Quoted 9 lines
Quoted 2 lines

Do you see another solution than renaming the _DLDIR_ version to the old name?

We either backport the patch as is and have the same functionality across the branches, or we do some special version of the patch for the LTS branches. In many ways I think it might be better to have things consistent rather than doing something special with all the confusion that can cause, even if people do have to actively adapt to the change.

Well, I'm not happy with a change of variables in existing releases. However the split would be complicated either way.

And as a bonus, I'm not able to reproduce on my systems (the lack of NFS might be a factor).

So, I'll say go for this backport even if unhappy with it.

Kind regards,
Marta

Quoted 2 lines
#10 lore raw

Re: [OE-core][scarthgap 1/9] cve_check: Use a local copy of the database during builds

Quoted 35 lines · Marta Rybczynska
Quoted 21 lines · On Wed, 30 Oct 2024, 07:40 Richard Purdie…
Quoted 2 lines · Marta Rybczynska

To make a summary, this patch did not solve the issue on master, it was the change of the database path that did.

Originally, I couldn't isolate the change to tell or not. Now, by changing the database path we've shown the issue is occurring on the branches without the patch.

Quoted 3 lines

Is this correct? And we still do not know what the cause is? So,
when we change the path on all branches, the issue is likely to come back ?

I think if we move the database operations to the local filesystem and out of DL_DIR, we're likely to stop seeing the issues. We know they're isolated to something about the older releases.

Quoted 4 lines

My problem with this patch is that it changes names of variables that are in use. It should at least not fail silently for people who decide to use a pinned database.

Quoted 8 lines
Quoted 2 lines

Do you see another solution than renaming the _DLDIR_ version to the old name?

We either backport the patch as is and have the same functionality across the branches, or we do some special version of the patch for the LTS branches. In many ways I think it might be better to have things consistent rather than doing something special with all the confusion that can cause, even if people do have to actively adapt to the change.

Quoted 5 lines

Well, I'm not happy with a change of variables in existing releases. However the split would be complicated either way.

And as a bonus, I'm not able to reproduce on my systems (the lack of NFS might be a factor).

So, I'll say go for this backport even if unhappy with it.

Thanks Marta, I will proceed with that.

Steve4