kernel August 2020

kernel@lists.fedoraproject.org

24 participants
137 discussions

[OS-BUILD PATCH] Fix branch creation for releases based on tags
by GitLab Bridge on behalf of jeremycline 10 May '21

10 May '21

From: Jeremy Cline <jcline(a)redhat.com> The release scripts are supposed create a set of branches if the release is based of a tag rather than a snapshot. davidlt noted on IRC that branches were missing for recent tags, namely rc3 and rc4. The reason is that the check for a signed tag expects $UPSTREAM_REF to be the tag. Resolve $UPSTREAM_REF to a tag (if it exists) prior to verifying the tag. Signed-off-by: Jeremy Cline <jcline(a)redhat.com> --- redhat/scripts/ci/ark-create-release.sh | 2 +- redhat/scripts/ci/ark-rebase-patches.sh | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/redhat/scripts/ci/ark-create-release.sh b/redhat/scripts/ci/ark-create-release.sh index fd57e533ddb4..83b102ceb54f 100755 --- a/redhat/scripts/ci/ark-create-release.sh +++ b/redhat/scripts/ci/ark-create-release.sh @@ -37,7 +37,7 @@ git checkout os-build git checkout os-build ./redhat/scripts/ci/ark-update-configs.sh "$UPSTREAM_REF" "$PROJECT_ID" -if git tag -v "$UPSTREAM_REF" > /dev/null 2>&1; then +if git tag -v "$(git describe --exact-match "$UPSTREAM_REF")"; then git checkout -b ark/"$UPSTREAM_REF" ark/patches/"$UPSTREAM_REF" RELEASE_BRANCHES=" ark/$UPSTREAM_REF ark/patches/$UPSTREAM_REF" else diff --git a/redhat/scripts/ci/ark-rebase-patches.sh b/redhat/scripts/ci/ark-rebase-patches.sh index 2ac22bcba3bd..4a02176bde88 100755 --- a/redhat/scripts/ci/ark-rebase-patches.sh +++ b/redhat/scripts/ci/ark-rebase-patches.sh @@ -102,7 +102,7 @@ else exit 2 fi -if git tag -v "$UPSTREAM_REF" > /dev/null 2>&1; then +if git tag -v "$(git describe --exact-match "$UPSTREAM_REF")"; then printf "Creating branch \"ark/patches/%s\"\n" "$UPSTREAM_REF" git branch ark/patches/"$UPSTREAM_REF" printf "Don't forget to run 'git push <remote> ark/patches/%s'\n" "$UPSTREAM_REF" -- 2.26.2

2 1

[OS-BUILD PATCH] [redhat] New configs in drivers/bus
by GitLab Bridge on behalf of jeremycline 06 May '21

06 May '21

6 7

Erase i686 from the kernel package's repository?
by Paul Bolle 22 Apr '21

22 Apr '21

Hi Justin, Fedora 31's release is imminent. This means that some time ago Fedora passed the point where (F31 and later) i686 releases were still possible, doesn't it? If so, could we please clean the kernel package's repository of all the things that are only relevant for i686? Because that would make my plans for the config generation [1] just a bit easier. Thanks, Paul Bolle [1] See https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.or…

4 16

[OS-BUILD PATCH] [redhat] New configs in drivers/gpu
by GitLab Bridge on behalf of jeremycline 16 Apr '21

16 Apr '21

From: Fedora Kernel Team <kernel-team(a)fedoraproject.org> Hi, As part of the ongoing rebase effort, the following configuration options need to be reviewed. As a reminder, the ARK configuration flow involves moving unreviewed configuration options from the pending directory to the ark directory. In the diff below, options are removed from the pending directory and added to the ark hierarchy. The final options that need to be ACKed are the files that are being added to the ark hierarchy. If the value for a file that is added should be changed, please reply with a better option. CONFIG_DRM_AMD_DC_DCN3_0: Choose this option if you want to have sienna_cichlid support for display engine Symbol: DRM_AMD_DC_DCN3_0 [=n] Type : bool Defined at drivers/gpu/drm/amd/display/Kconfig:20 Prompt: DCN 3.0 family Depends on: HAS_IOMEM [=y] && DRM [=m] && DRM_AMDGPU [=m] && DRM_AMD_DC [=y] && X86 [=y] && DRM_AMD_DC_DCN [=y] Location: -> Device Drivers -> Graphics support -> AMD GPU (DRM_AMDGPU [=m]) -> Display Engine Configuration --- CONFIG_NOUVEAU_DEBUG_PUSH: Say Y here if you want to enable verbose push buffer debug output and sanity checks. Symbol: NOUVEAU_DEBUG_PUSH [=n] Type : bool Defined at drivers/gpu/drm/nouveau/Kconfig:79 Prompt: Enable additional push buffer debugging Depends on: HAS_IOMEM [=y] && DRM_NOUVEAU [=m] Location: -> Device Drivers -> Graphics support -> Nouveau (NVIDIA) cards (DRM_NOUVEAU [=m]) --- Cc: David Airlie <airlied(a)redhat.com> Cc: Adam Jackson <ajax(a)redhat.com> Signed-off-by: Fedora Kernel Team <kernel-team(a)fedoraproject.org> --- .../common/generic/CONFIG_DRM_AMD_DC_DCN3_0 | 1 + .../common/generic/CONFIG_NOUVEAU_DEBUG_PUSH | 1 + .../generic/CONFIG_DRM_AMD_DC_DCN3_0 | 19 ------------------- .../generic/CONFIG_NOUVEAU_DEBUG_PUSH | 18 ------------------ 4 files changed, 2 insertions(+), 37 deletions(-) create mode 100644 redhat/configs/common/generic/CONFIG_DRM_AMD_DC_DCN3_0 create mode 100644 redhat/configs/common/generic/CONFIG_NOUVEAU_DEBUG_PUSH delete mode 100644 redhat/configs/pending-common/generic/CONFIG_DRM_AMD_DC_DCN3_0 delete mode 100644 redhat/configs/pending-common/generic/CONFIG_NOUVEAU_DEBUG_PUSH diff --git a/redhat/configs/common/generic/CONFIG_DRM_AMD_DC_DCN3_0 b/redhat/configs/common/generic/CONFIG_DRM_AMD_DC_DCN3_0 new file mode 100644 index 000000000000..708bd22e809e --- /dev/null +++ b/redhat/configs/common/generic/CONFIG_DRM_AMD_DC_DCN3_0 @@ -0,0 +1 @@ +# CONFIG_DRM_AMD_DC_DCN3_0 is not set diff --git a/redhat/configs/common/generic/CONFIG_NOUVEAU_DEBUG_PUSH b/redhat/configs/common/generic/CONFIG_NOUVEAU_DEBUG_PUSH new file mode 100644 index 000000000000..5614c2c49809 --- /dev/null +++ b/redhat/configs/common/generic/CONFIG_NOUVEAU_DEBUG_PUSH @@ -0,0 +1 @@ +# CONFIG_NOUVEAU_DEBUG_PUSH is not set diff --git a/redhat/configs/pending-common/generic/CONFIG_DRM_AMD_DC_DCN3_0 b/redhat/configs/pending-common/generic/CONFIG_DRM_AMD_DC_DCN3_0 deleted file mode 100644 index f6a3c764772e..000000000000 --- a/redhat/configs/pending-common/generic/CONFIG_DRM_AMD_DC_DCN3_0 +++ /dev/null @@ -1,19 +0,0 @@ -# CONFIG_DRM_AMD_DC_DCN3_0: -# -# Choose this option if you want to have -# sienna_cichlid support for display engine -# -# Symbol: DRM_AMD_DC_DCN3_0 [=n] -# Type : bool -# Defined at drivers/gpu/drm/amd/display/Kconfig:20 -# Prompt: DCN 3.0 family -# Depends on: HAS_IOMEM [=y] && DRM [=m] && DRM_AMDGPU [=m] && DRM_AMD_DC [=y] && X86 [=y] && DRM_AMD_DC_DCN [=y] -# Location: -# -> Device Drivers -# -> Graphics support -# -> AMD GPU (DRM_AMDGPU [=m]) -# -> Display Engine Configuration -# -# -# -# CONFIG_DRM_AMD_DC_DCN3_0 is not set diff --git a/redhat/configs/pending-common/generic/CONFIG_NOUVEAU_DEBUG_PUSH b/redhat/configs/pending-common/generic/CONFIG_NOUVEAU_DEBUG_PUSH deleted file mode 100644 index c28eaf703375..000000000000 --- a/redhat/configs/pending-common/generic/CONFIG_NOUVEAU_DEBUG_PUSH +++ /dev/null @@ -1,18 +0,0 @@ -# CONFIG_NOUVEAU_DEBUG_PUSH: -# -# Say Y here if you want to enable verbose push buffer debug output -# and sanity checks. -# -# Symbol: NOUVEAU_DEBUG_PUSH [=n] -# Type : bool -# Defined at drivers/gpu/drm/nouveau/Kconfig:79 -# Prompt: Enable additional push buffer debugging -# Depends on: HAS_IOMEM [=y] && DRM_NOUVEAU [=m] -# Location: -# -> Device Drivers -# -> Graphics support -# -> Nouveau (NVIDIA) cards (DRM_NOUVEAU [=m]) -# -# -# -# CONFIG_NOUVEAU_DEBUG_PUSH is not set -- GitLab

7 8

[OS-BUILD PATCH] [redhat] New configs in drivers/power
by GitLab Bridge on behalf of jeremycline 14 Apr '21

14 Apr '21

3 3

[OS-BUILD PATCH] redhat/Makefile.common: Fix 'fatal: Not a valid object name master' error
by GitLab Bridge on behalf of prarit 06 Apr '21

06 Apr '21

From: Prarit Bhargava <prarit(a)redhat.com> Hey everyone, I spent 5-10 mins looking at the 'fatal: Not a valid object name master' error this morning. It comes from the redhat/Makefile.common:45 MERGE_BASE:=$(shell git merge-base HEAD master) There is no master branch so 'git merge-base HEAD master' returns an error with message 'fatal: Not a valid object name master'. If I add a Makefile target to dump the MERGE_BASE, TAG, and SNAPSHOT variables I get MERGE_BASE= TAG=v5.8-rc7-1028-g305981a20220 SNAPSHOT=1 I thought about changing the MERGE_BASE value to MERGE_BASE:=$(shell git merge-base HEAD master >& /dev/null && git merge-base HEAD master || git merge-base HEAD upstream/master) but that seems really clunky. Anyone have a better idea? Maybe it should only be MERGE_BASE:$(git merge-base HEAD upstream/master) ? In any case, after changing the MERGE_BASE line in Makefile.common, and dumping the values the 'fatal' error no longer occurs and MERGE_BASE is set correctly: MERGE_BASE=d3590ebf6f91350192737dd1d1b219c05277f067 TAG=v5.8-rc7-39-gd3590ebf6f91 SNAPSHOT=1 P. ---8<---- Executing some redhat/Makefile targets results in a 'fatal: Not a valid object name master' error. This occurs because MERGE_BASE is calculated against the master branch which may not exist. Check to see if master branch exists and if it doesn't use upstream/master instead. Signed-off-by: Prarit Bhargava <prarit(a)redhat.com> --- redhat/Makefile.common | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/redhat/Makefile.common b/redhat/Makefile.common index 6d048415ba6d..e67f2c9dec50 100644 --- a/redhat/Makefile.common +++ b/redhat/Makefile.common @@ -41,8 +41,7 @@ else VERSION_ON_UPSTREAM:=0 endif ifeq ($(VERSION_ON_UPSTREAM),1) - # master is expected to track mainline. - MERGE_BASE:=$(shell git merge-base HEAD master) + MERGE_BASE:=$(shell git merge-base HEAD master >& /dev/null && git merge-base HEAD master || git merge-base HEAD upstream/master) TAG:=$(shell git describe $(MERGE_BASE)) # a snapshot off of a tagged git is of the form [tag]-[cnt]-g[hash] SNAPSHOT:=$(shell echo $(TAG) | grep -c '\-g') -- GitLab

3 4

[OS-BUILD PATCH] [redhat] New configs in fs/Kconfig
by GitLab Bridge on behalf of jeremycline 29 Mar '21

29 Mar '21

From: Fedora Kernel Team <kernel-team(a)fedoraproject.org> Hi, As part of the ongoing rebase effort, the following configuration options need to be reviewed. As a reminder, the ARK configuration flow involves moving unreviewed configuration options from the pending directory to the ark directory. In the diff below, options are removed from the pending directory and added to the ark hierarchy. The final options that need to be ACKed are the files that are being added to the ark hierarchy. If the value for a file that is added should be changed, please reply with a better option. CONFIG_TMPFS_INODE64: tmpfs has historically used only inode numbers as wide as an unsigned int. In some cases this can cause wraparound, potentially resulting in multiple files with the same inode number on a single device. This option makes tmpfs use the full width of ino_t by default, without needing to specify the inode64 option when mounting. But if a long-lived tmpfs is to be accessed by 32-bit applications so ancient that opening a file larger than 2GiB fails with EINVAL, then the INODE64 config option and inode64 mount option risk operations failing with EOVERFLOW once 33-bit inode numbers are reached. To override this configured default, use the inode32 or inode64 option when mounting. If unsure, say N. Symbol: TMPFS_INODE64 [=n] Type : bool Defined at fs/Kconfig:204 Prompt: Use 64-bit ino_t by default in tmpfs Depends on: TMPFS [=y] && 64BIT [=y] Location: -> File systems -> Pseudo filesystems -> Tmpfs virtual memory file system support (former shm fs) (TMPFS [=y]) --- Signed-off-by: Fedora Kernel Team <kernel-team(a)fedoraproject.org> --- .../common/generic/CONFIG_TMPFS_INODE64 | 1 + .../generic/CONFIG_TMPFS_INODE64 | 31 ------------------- 2 files changed, 1 insertion(+), 31 deletions(-) create mode 100644 redhat/configs/common/generic/CONFIG_TMPFS_INODE64 delete mode 100644 redhat/configs/pending-common/generic/CONFIG_TMPFS_INODE64 diff --git a/redhat/configs/common/generic/CONFIG_TMPFS_INODE64 b/redhat/configs/common/generic/CONFIG_TMPFS_INODE64 new file mode 100644 index 000000000000..cce8a2274ff4 --- /dev/null +++ b/redhat/configs/common/generic/CONFIG_TMPFS_INODE64 @@ -0,0 +1 @@ +# CONFIG_TMPFS_INODE64 is not set diff --git a/redhat/configs/pending-common/generic/CONFIG_TMPFS_INODE64 b/redhat/configs/pending-common/generic/CONFIG_TMPFS_INODE64 deleted file mode 100644 index a1c19f28cc41..000000000000 --- a/redhat/configs/pending-common/generic/CONFIG_TMPFS_INODE64 +++ /dev/null @@ -1,31 +0,0 @@ -# CONFIG_TMPFS_INODE64: -# -# tmpfs has historically used only inode numbers as wide as an unsigned -# int. In some cases this can cause wraparound, potentially resulting -# in multiple files with the same inode number on a single device. This -# option makes tmpfs use the full width of ino_t by default, without -# needing to specify the inode64 option when mounting. -# -# But if a long-lived tmpfs is to be accessed by 32-bit applications so -# ancient that opening a file larger than 2GiB fails with EINVAL, then -# the INODE64 config option and inode64 mount option risk operations -# failing with EOVERFLOW once 33-bit inode numbers are reached. -# -# To override this configured default, use the inode32 or inode64 -# option when mounting. -# -# If unsure, say N. -# -# Symbol: TMPFS_INODE64 [=n] -# Type : bool -# Defined at fs/Kconfig:204 -# Prompt: Use 64-bit ino_t by default in tmpfs -# Depends on: TMPFS [=y] && 64BIT [=y] -# Location: -# -> File systems -# -> Pseudo filesystems -# -> Tmpfs virtual memory file system support (former shm fs) (TMPFS [=y]) -# -# -# -# CONFIG_TMPFS_INODE64 is not set -- GitLab

4 3

[OS-BUILD PATCH] [redhat] New configs in arch/arm64
by GitLab Bridge on behalf of jeremycline 29 Mar '21

29 Mar '21

From: "CKI@GitLab" <cki-project(a)redhat.com> Hi, As part of the ongoing rebase effort, the following configuration options need to be reviewed. As a reminder, the ARK configuration flow involves moving unreviewed configuration options from the pending directory to the ark directory. In the diff below, options are removed from the pending directory and added to the ark hierarchy. The final options that need to be ACKed are the files that are being added to the ark hierarchy. If the value for a file that is added should be changed, please reply with a better option. CONFIG_ARM64_BTI: Branch Target Identification (part of the ARMv8.5 Extensions) provides a mechanism to limit the set of locations to which computed branch instructions such as BR or BLR can jump. To make use of BTI on CPUs that support it, say Y. BTI is intended to provide complementary protection to other control flow integrity protection mechanisms, such as the Pointer authentication mechanism provided as part of the ARMv8.3 Extensions. For this reason, it does not make sense to enable this option without also enabling support for pointer authentication. Thus, when enabling this option you should also select ARM64_PTR_AUTH=y. Userspace binaries must also be specifically compiled to make use of this mechanism. If you say N here or the hardware does not support BTI, such binaries can still run, but you get no additional enforcement of branch destinations. Symbol: ARM64_BTI [=y] Type : bool Defined at arch/arm64/Kconfig:1594 Prompt: Branch Target Identification support Location: -> Kernel Features -> ARMv8.5 architectural features --- Cc: Mark Salter <msalter(a)redhat.com> Signed-off-by: CKI@GitLab <cki-project(a)redhat.com> --- .../configs/common/generic/CONFIG_ARM64_BTI | 1 + .../pending-common/generic/CONFIG_ARM64_BTI | 31 ------------------- 2 files changed, 1 insertion(+), 31 deletions(-) create mode 100644 redhat/configs/common/generic/CONFIG_ARM64_BTI delete mode 100644 redhat/configs/pending-common/generic/CONFIG_ARM64_BTI diff --git a/redhat/configs/common/generic/CONFIG_ARM64_BTI b/redhat/configs/common/generic/CONFIG_ARM64_BTI new file mode 100644 index 000000000000..fb0274de0d49 --- /dev/null +++ b/redhat/configs/common/generic/CONFIG_ARM64_BTI @@ -0,0 +1 @@ +CONFIG_ARM64_BTI=y diff --git a/redhat/configs/pending-common/generic/CONFIG_ARM64_BTI b/redhat/configs/pending-common/generic/CONFIG_ARM64_BTI deleted file mode 100644 index 5af4d535b648..000000000000 --- a/redhat/configs/pending-common/generic/CONFIG_ARM64_BTI +++ /dev/null @@ -1,31 +0,0 @@ -# CONFIG_ARM64_BTI: -# -# Branch Target Identification (part of the ARMv8.5 Extensions) -# provides a mechanism to limit the set of locations to which computed -# branch instructions such as BR or BLR can jump. -# -# To make use of BTI on CPUs that support it, say Y. -# -# BTI is intended to provide complementary protection to other control -# flow integrity protection mechanisms, such as the Pointer -# authentication mechanism provided as part of the ARMv8.3 Extensions. -# For this reason, it does not make sense to enable this option without -# also enabling support for pointer authentication. Thus, when -# enabling this option you should also select ARM64_PTR_AUTH=y. -# -# Userspace binaries must also be specifically compiled to make use of -# this mechanism. If you say N here or the hardware does not support -# BTI, such binaries can still run, but you get no additional -# enforcement of branch destinations. -# -# Symbol: ARM64_BTI [=y] -# Type : bool -# Defined at arch/arm64/Kconfig:1594 -# Prompt: Branch Target Identification support -# Location: -# -> Kernel Features -# -> ARMv8.5 architectural features -# -# -# -CONFIG_ARM64_BTI=y -- 2.26.2

4 3

[OS-BUILD PATCH] [redhat] New configs in drivers/watchdog
by GitLab Bridge on behalf of jeremycline 29 Mar '21

29 Mar '21

3 2

[OS-BUILD PATCH] [redhat] New configs in arch/powerpc
by GitLab Bridge on behalf of jeremycline 24 Mar '21

24 Mar '21

5 4

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

kernel August 2020