From: Sasha Levin Date: Fri, 23 May 2025 12:18:00 +0000 (-0400) Subject: Drop exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch X-Git-Tag: v6.12.31~91 X-Git-Url: http://git.ipfire.org/?a=commitdiff_plain;h=1df32f2803dbf55648055400f5bb5579f6aa4bcf;p=thirdparty%2Fkernel%2Fstable-queue.git Drop exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch Signed-off-by: Sasha Levin --- diff --git a/queue-5.10/exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch b/queue-5.10/exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch deleted file mode 100644 index 7cd5560230..0000000000 --- a/queue-5.10/exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch +++ /dev/null @@ -1,52 +0,0 @@ -From d9daeb4c55dd4ca79f66d7798c3f23045d3053df Mon Sep 17 00:00:00 2001 -From: Sasha Levin -Date: Mon, 24 Mar 2025 18:19:41 +0100 -Subject: exit: fix the usage of delay_group_leader->exit_code in - do_notify_parent() and pidfs_exit() - -From: Oleg Nesterov - -[ Upstream commit 9133607de37a4887c6f89ed937176a0a0c1ebb17 ] - -Consider a process with a group leader L and a sub-thread T. -L does sys_exit(1), then T does sys_exit_group(2). - -In this case wait_task_zombie(L) will notice SIGNAL_GROUP_EXIT and use -L->signal->group_exit_code, this is correct. - -But, before that, do_notify_parent(L) called by release_task(T) will use -L->exit_code != L->signal->group_exit_code, and this is not consistent. -We don't really care, I think that nobody relies on the info which comes -with SIGCHLD, if nothing else SIGCHLD < SIGRTMIN can be queued only once. - -But pidfs_exit() is more problematic, I think pidfs_exit_info->exit_code -should report ->group_exit_code in this case, just like wait_task_zombie(). - -TODO: with this change we can hopefully cleanup (or may be even kill) the -similar SIGNAL_GROUP_EXIT checks, at least in wait_task_zombie(). - -Signed-off-by: Oleg Nesterov -Link: https://lore.kernel.org/r/20250324171941.GA13114@redhat.com -Signed-off-by: Christian Brauner -Signed-off-by: Sasha Levin ---- - kernel/exit.c | 3 +++ - 1 file changed, 3 insertions(+) - -diff --git a/kernel/exit.c b/kernel/exit.c -index af9c8e794e4d7..05f682cfdd6a7 100644 ---- a/kernel/exit.c -+++ b/kernel/exit.c -@@ -254,6 +254,9 @@ void release_task(struct task_struct *p) - leader = p->group_leader; - if (leader != p && thread_group_empty(leader) - && leader->exit_state == EXIT_ZOMBIE) { -+ /* for pidfs_exit() and do_notify_parent() */ -+ if (leader->signal->flags & SIGNAL_GROUP_EXIT) -+ leader->exit_code = leader->signal->group_exit_code; - /* - * If we were the last child thread and the leader has - * exited already, and the leader's parent ignores SIGCHLD, --- -2.39.5 - diff --git a/queue-5.10/series b/queue-5.10/series index 6e403cd78d..2243e76152 100644 --- a/queue-5.10/series +++ b/queue-5.10/series @@ -126,7 +126,6 @@ nfsv4-treat-enetunreach-errors-as-fatal-for-state-re.patch sunrpc-rpc_clnt_set_transport-must-not-change-the-au.patch sunrpc-rpcbind-should-never-reset-the-port-to-the-va.patch thermal-drivers-qoriq-power-down-tmu-on-system-suspe.patch -exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch dql-fix-dql-limit-value-when-reset.patch tools-build-don-t-pass-test-log-files-to-linker.patch pnfs-flexfiles-report-enetdown-as-a-connection-error.patch diff --git a/queue-5.15/exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch b/queue-5.15/exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch deleted file mode 100644 index 9dafba2806..0000000000 --- a/queue-5.15/exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch +++ /dev/null @@ -1,52 +0,0 @@ -From f74f25150b68253a59dcab73d960028fc6992f34 Mon Sep 17 00:00:00 2001 -From: Sasha Levin -Date: Mon, 24 Mar 2025 18:19:41 +0100 -Subject: exit: fix the usage of delay_group_leader->exit_code in - do_notify_parent() and pidfs_exit() - -From: Oleg Nesterov - -[ Upstream commit 9133607de37a4887c6f89ed937176a0a0c1ebb17 ] - -Consider a process with a group leader L and a sub-thread T. -L does sys_exit(1), then T does sys_exit_group(2). - -In this case wait_task_zombie(L) will notice SIGNAL_GROUP_EXIT and use -L->signal->group_exit_code, this is correct. - -But, before that, do_notify_parent(L) called by release_task(T) will use -L->exit_code != L->signal->group_exit_code, and this is not consistent. -We don't really care, I think that nobody relies on the info which comes -with SIGCHLD, if nothing else SIGCHLD < SIGRTMIN can be queued only once. - -But pidfs_exit() is more problematic, I think pidfs_exit_info->exit_code -should report ->group_exit_code in this case, just like wait_task_zombie(). - -TODO: with this change we can hopefully cleanup (or may be even kill) the -similar SIGNAL_GROUP_EXIT checks, at least in wait_task_zombie(). - -Signed-off-by: Oleg Nesterov -Link: https://lore.kernel.org/r/20250324171941.GA13114@redhat.com -Signed-off-by: Christian Brauner -Signed-off-by: Sasha Levin ---- - kernel/exit.c | 3 +++ - 1 file changed, 3 insertions(+) - -diff --git a/kernel/exit.c b/kernel/exit.c -index 890e5cb6799b0..04fc65f2b690d 100644 ---- a/kernel/exit.c -+++ b/kernel/exit.c -@@ -254,6 +254,9 @@ void release_task(struct task_struct *p) - leader = p->group_leader; - if (leader != p && thread_group_empty(leader) - && leader->exit_state == EXIT_ZOMBIE) { -+ /* for pidfs_exit() and do_notify_parent() */ -+ if (leader->signal->flags & SIGNAL_GROUP_EXIT) -+ leader->exit_code = leader->signal->group_exit_code; - /* - * If we were the last child thread and the leader has - * exited already, and the leader's parent ignores SIGCHLD, --- -2.39.5 - diff --git a/queue-5.15/series b/queue-5.15/series index 2f7bd9292c..45d17e925d 100644 --- a/queue-5.15/series +++ b/queue-5.15/series @@ -16,7 +16,6 @@ nfsv4-treat-enetunreach-errors-as-fatal-for-state-re.patch sunrpc-rpc_clnt_set_transport-must-not-change-the-au.patch sunrpc-rpcbind-should-never-reset-the-port-to-the-va.patch thermal-drivers-qoriq-power-down-tmu-on-system-suspe.patch -exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch dql-fix-dql-limit-value-when-reset.patch lockdep-fix-wait-context-check-on-softirq-for-preemp.patch pci-dwc-ep-ensure-proper-iteration-over-outbound-map.patch diff --git a/queue-5.4/exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch b/queue-5.4/exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch deleted file mode 100644 index 95b0ef0d8f..0000000000 --- a/queue-5.4/exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch +++ /dev/null @@ -1,52 +0,0 @@ -From af236fb6e7c5cd027b37ec565c28f3758dac1b54 Mon Sep 17 00:00:00 2001 -From: Sasha Levin -Date: Mon, 24 Mar 2025 18:19:41 +0100 -Subject: exit: fix the usage of delay_group_leader->exit_code in - do_notify_parent() and pidfs_exit() - -From: Oleg Nesterov - -[ Upstream commit 9133607de37a4887c6f89ed937176a0a0c1ebb17 ] - -Consider a process with a group leader L and a sub-thread T. -L does sys_exit(1), then T does sys_exit_group(2). - -In this case wait_task_zombie(L) will notice SIGNAL_GROUP_EXIT and use -L->signal->group_exit_code, this is correct. - -But, before that, do_notify_parent(L) called by release_task(T) will use -L->exit_code != L->signal->group_exit_code, and this is not consistent. -We don't really care, I think that nobody relies on the info which comes -with SIGCHLD, if nothing else SIGCHLD < SIGRTMIN can be queued only once. - -But pidfs_exit() is more problematic, I think pidfs_exit_info->exit_code -should report ->group_exit_code in this case, just like wait_task_zombie(). - -TODO: with this change we can hopefully cleanup (or may be even kill) the -similar SIGNAL_GROUP_EXIT checks, at least in wait_task_zombie(). - -Signed-off-by: Oleg Nesterov -Link: https://lore.kernel.org/r/20250324171941.GA13114@redhat.com -Signed-off-by: Christian Brauner -Signed-off-by: Sasha Levin ---- - kernel/exit.c | 3 +++ - 1 file changed, 3 insertions(+) - -diff --git a/kernel/exit.c b/kernel/exit.c -index 56d3a099825fb..5015ecdda6d95 100644 ---- a/kernel/exit.c -+++ b/kernel/exit.c -@@ -262,6 +262,9 @@ void release_task(struct task_struct *p) - leader = p->group_leader; - if (leader != p && thread_group_empty(leader) - && leader->exit_state == EXIT_ZOMBIE) { -+ /* for pidfs_exit() and do_notify_parent() */ -+ if (leader->signal->flags & SIGNAL_GROUP_EXIT) -+ leader->exit_code = leader->signal->group_exit_code; - /* - * If we were the last child thread and the leader has - * exited already, and the leader's parent ignores SIGCHLD, --- -2.39.5 - diff --git a/queue-5.4/series b/queue-5.4/series index 58c3b62e2f..b0909da1ca 100644 --- a/queue-5.4/series +++ b/queue-5.4/series @@ -92,7 +92,6 @@ fbdev-fsl-diu-fb-add-missing-device_remove_file.patch fbdev-core-tileblit-implement-missing-margin-clearin.patch nfsv4-treat-enetunreach-errors-as-fatal-for-state-re.patch sunrpc-rpc_clnt_set_transport-must-not-change-the-au.patch -exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch dql-fix-dql-limit-value-when-reset.patch tools-build-don-t-pass-test-log-files-to-linker.patch pnfs-flexfiles-report-enetdown-as-a-connection-error.patch diff --git a/queue-6.1/exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch b/queue-6.1/exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch deleted file mode 100644 index bdcd0576f0..0000000000 --- a/queue-6.1/exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch +++ /dev/null @@ -1,52 +0,0 @@ -From a811bf878f15ee50b7459e4019c733e4788be371 Mon Sep 17 00:00:00 2001 -From: Sasha Levin -Date: Mon, 24 Mar 2025 18:19:41 +0100 -Subject: exit: fix the usage of delay_group_leader->exit_code in - do_notify_parent() and pidfs_exit() - -From: Oleg Nesterov - -[ Upstream commit 9133607de37a4887c6f89ed937176a0a0c1ebb17 ] - -Consider a process with a group leader L and a sub-thread T. -L does sys_exit(1), then T does sys_exit_group(2). - -In this case wait_task_zombie(L) will notice SIGNAL_GROUP_EXIT and use -L->signal->group_exit_code, this is correct. - -But, before that, do_notify_parent(L) called by release_task(T) will use -L->exit_code != L->signal->group_exit_code, and this is not consistent. -We don't really care, I think that nobody relies on the info which comes -with SIGCHLD, if nothing else SIGCHLD < SIGRTMIN can be queued only once. - -But pidfs_exit() is more problematic, I think pidfs_exit_info->exit_code -should report ->group_exit_code in this case, just like wait_task_zombie(). - -TODO: with this change we can hopefully cleanup (or may be even kill) the -similar SIGNAL_GROUP_EXIT checks, at least in wait_task_zombie(). - -Signed-off-by: Oleg Nesterov -Link: https://lore.kernel.org/r/20250324171941.GA13114@redhat.com -Signed-off-by: Christian Brauner -Signed-off-by: Sasha Levin ---- - kernel/exit.c | 3 +++ - 1 file changed, 3 insertions(+) - -diff --git a/kernel/exit.c b/kernel/exit.c -index 156283b3c1bf6..49e4792405008 100644 ---- a/kernel/exit.c -+++ b/kernel/exit.c -@@ -263,6 +263,9 @@ void release_task(struct task_struct *p) - leader = p->group_leader; - if (leader != p && thread_group_empty(leader) - && leader->exit_state == EXIT_ZOMBIE) { -+ /* for pidfs_exit() and do_notify_parent() */ -+ if (leader->signal->flags & SIGNAL_GROUP_EXIT) -+ leader->exit_code = leader->signal->group_exit_code; - /* - * If we were the last child thread and the leader has - * exited already, and the leader's parent ignores SIGCHLD, --- -2.39.5 - diff --git a/queue-6.1/series b/queue-6.1/series index 000b4ad5d1..740b85ec1a 100644 --- a/queue-6.1/series +++ b/queue-6.1/series @@ -36,7 +36,6 @@ nfsv4-treat-enetunreach-errors-as-fatal-for-state-re.patch sunrpc-rpc_clnt_set_transport-must-not-change-the-au.patch sunrpc-rpcbind-should-never-reset-the-port-to-the-va.patch thermal-drivers-qoriq-power-down-tmu-on-system-suspe.patch -exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch dql-fix-dql-limit-value-when-reset.patch lockdep-fix-wait-context-check-on-softirq-for-preemp.patch objtool-properly-disable-uaccess-validation.patch diff --git a/queue-6.12/exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch b/queue-6.12/exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch deleted file mode 100644 index d840b6ff36..0000000000 --- a/queue-6.12/exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch +++ /dev/null @@ -1,52 +0,0 @@ -From 9591f9ea1e2a52280934f3c69e9687a87b93e360 Mon Sep 17 00:00:00 2001 -From: Sasha Levin -Date: Mon, 24 Mar 2025 18:19:41 +0100 -Subject: exit: fix the usage of delay_group_leader->exit_code in - do_notify_parent() and pidfs_exit() - -From: Oleg Nesterov - -[ Upstream commit 9133607de37a4887c6f89ed937176a0a0c1ebb17 ] - -Consider a process with a group leader L and a sub-thread T. -L does sys_exit(1), then T does sys_exit_group(2). - -In this case wait_task_zombie(L) will notice SIGNAL_GROUP_EXIT and use -L->signal->group_exit_code, this is correct. - -But, before that, do_notify_parent(L) called by release_task(T) will use -L->exit_code != L->signal->group_exit_code, and this is not consistent. -We don't really care, I think that nobody relies on the info which comes -with SIGCHLD, if nothing else SIGCHLD < SIGRTMIN can be queued only once. - -But pidfs_exit() is more problematic, I think pidfs_exit_info->exit_code -should report ->group_exit_code in this case, just like wait_task_zombie(). - -TODO: with this change we can hopefully cleanup (or may be even kill) the -similar SIGNAL_GROUP_EXIT checks, at least in wait_task_zombie(). - -Signed-off-by: Oleg Nesterov -Link: https://lore.kernel.org/r/20250324171941.GA13114@redhat.com -Signed-off-by: Christian Brauner -Signed-off-by: Sasha Levin ---- - kernel/exit.c | 3 +++ - 1 file changed, 3 insertions(+) - -diff --git a/kernel/exit.c b/kernel/exit.c -index 619f0014c33be..9d888f26bf784 100644 ---- a/kernel/exit.c -+++ b/kernel/exit.c -@@ -266,6 +266,9 @@ void release_task(struct task_struct *p) - leader = p->group_leader; - if (leader != p && thread_group_empty(leader) - && leader->exit_state == EXIT_ZOMBIE) { -+ /* for pidfs_exit() and do_notify_parent() */ -+ if (leader->signal->flags & SIGNAL_GROUP_EXIT) -+ leader->exit_code = leader->signal->group_exit_code; - /* - * If we were the last child thread and the leader has - * exited already, and the leader's parent ignores SIGCHLD, --- -2.39.5 - diff --git a/queue-6.12/series b/queue-6.12/series index 7afdb13c14..45b6dab0dc 100644 --- a/queue-6.12/series +++ b/queue-6.12/series @@ -81,7 +81,6 @@ thermal-drivers-mediatek-lvts-start-sensor-interrupt.patch thermal-drivers-qoriq-power-down-tmu-on-system-suspe.patch bluetooth-btmtksdio-prevent-enabling-interrupts-afte.patch bluetooth-disable-sco-support-if-read_voice_setting-.patch -exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch dql-fix-dql-limit-value-when-reset.patch lockdep-fix-wait-context-check-on-softirq-for-preemp.patch objtool-properly-disable-uaccess-validation.patch diff --git a/queue-6.14/exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch b/queue-6.14/exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch deleted file mode 100644 index ae6aac555a..0000000000 --- a/queue-6.14/exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch +++ /dev/null @@ -1,52 +0,0 @@ -From 909fe6c95cb78648e9b6398dbb8c3bba92459c46 Mon Sep 17 00:00:00 2001 -From: Sasha Levin -Date: Mon, 24 Mar 2025 18:19:41 +0100 -Subject: exit: fix the usage of delay_group_leader->exit_code in - do_notify_parent() and pidfs_exit() - -From: Oleg Nesterov - -[ Upstream commit 9133607de37a4887c6f89ed937176a0a0c1ebb17 ] - -Consider a process with a group leader L and a sub-thread T. -L does sys_exit(1), then T does sys_exit_group(2). - -In this case wait_task_zombie(L) will notice SIGNAL_GROUP_EXIT and use -L->signal->group_exit_code, this is correct. - -But, before that, do_notify_parent(L) called by release_task(T) will use -L->exit_code != L->signal->group_exit_code, and this is not consistent. -We don't really care, I think that nobody relies on the info which comes -with SIGCHLD, if nothing else SIGCHLD < SIGRTMIN can be queued only once. - -But pidfs_exit() is more problematic, I think pidfs_exit_info->exit_code -should report ->group_exit_code in this case, just like wait_task_zombie(). - -TODO: with this change we can hopefully cleanup (or may be even kill) the -similar SIGNAL_GROUP_EXIT checks, at least in wait_task_zombie(). - -Signed-off-by: Oleg Nesterov -Link: https://lore.kernel.org/r/20250324171941.GA13114@redhat.com -Signed-off-by: Christian Brauner -Signed-off-by: Sasha Levin ---- - kernel/exit.c | 3 +++ - 1 file changed, 3 insertions(+) - -diff --git a/kernel/exit.c b/kernel/exit.c -index 3485e5fc499e4..6bb59b16e33e1 100644 ---- a/kernel/exit.c -+++ b/kernel/exit.c -@@ -265,6 +265,9 @@ void release_task(struct task_struct *p) - leader = p->group_leader; - if (leader != p && thread_group_empty(leader) - && leader->exit_state == EXIT_ZOMBIE) { -+ /* for pidfs_exit() and do_notify_parent() */ -+ if (leader->signal->flags & SIGNAL_GROUP_EXIT) -+ leader->exit_code = leader->signal->group_exit_code; - /* - * If we were the last child thread and the leader has - * exited already, and the leader's parent ignores SIGCHLD, --- -2.39.5 - diff --git a/queue-6.14/series b/queue-6.14/series index 7676e55c5a..0f2d41e895 100644 --- a/queue-6.14/series +++ b/queue-6.14/series @@ -85,7 +85,6 @@ thermal-drivers-mediatek-lvts-start-sensor-interrupt.patch thermal-drivers-qoriq-power-down-tmu-on-system-suspe.patch bluetooth-btmtksdio-prevent-enabling-interrupts-afte.patch bluetooth-disable-sco-support-if-read_voice_setting-.patch -exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch risc-v-add-vector-extension-validation-checks.patch dql-fix-dql-limit-value-when-reset.patch lockdep-fix-wait-context-check-on-softirq-for-preemp.patch diff --git a/queue-6.6/exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch b/queue-6.6/exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch deleted file mode 100644 index 8981014a2d..0000000000 --- a/queue-6.6/exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch +++ /dev/null @@ -1,52 +0,0 @@ -From a519a9b68e0f8ab06feb6ee7e4514e2b36a1c0d0 Mon Sep 17 00:00:00 2001 -From: Sasha Levin -Date: Mon, 24 Mar 2025 18:19:41 +0100 -Subject: exit: fix the usage of delay_group_leader->exit_code in - do_notify_parent() and pidfs_exit() - -From: Oleg Nesterov - -[ Upstream commit 9133607de37a4887c6f89ed937176a0a0c1ebb17 ] - -Consider a process with a group leader L and a sub-thread T. -L does sys_exit(1), then T does sys_exit_group(2). - -In this case wait_task_zombie(L) will notice SIGNAL_GROUP_EXIT and use -L->signal->group_exit_code, this is correct. - -But, before that, do_notify_parent(L) called by release_task(T) will use -L->exit_code != L->signal->group_exit_code, and this is not consistent. -We don't really care, I think that nobody relies on the info which comes -with SIGCHLD, if nothing else SIGCHLD < SIGRTMIN can be queued only once. - -But pidfs_exit() is more problematic, I think pidfs_exit_info->exit_code -should report ->group_exit_code in this case, just like wait_task_zombie(). - -TODO: with this change we can hopefully cleanup (or may be even kill) the -similar SIGNAL_GROUP_EXIT checks, at least in wait_task_zombie(). - -Signed-off-by: Oleg Nesterov -Link: https://lore.kernel.org/r/20250324171941.GA13114@redhat.com -Signed-off-by: Christian Brauner -Signed-off-by: Sasha Levin ---- - kernel/exit.c | 3 +++ - 1 file changed, 3 insertions(+) - -diff --git a/kernel/exit.c b/kernel/exit.c -index 3540b2c9b1b6a..1b7257c12cb10 100644 ---- a/kernel/exit.c -+++ b/kernel/exit.c -@@ -264,6 +264,9 @@ void release_task(struct task_struct *p) - leader = p->group_leader; - if (leader != p && thread_group_empty(leader) - && leader->exit_state == EXIT_ZOMBIE) { -+ /* for pidfs_exit() and do_notify_parent() */ -+ if (leader->signal->flags & SIGNAL_GROUP_EXIT) -+ leader->exit_code = leader->signal->group_exit_code; - /* - * If we were the last child thread and the leader has - * exited already, and the leader's parent ignores SIGCHLD, --- -2.39.5 - diff --git a/queue-6.6/series b/queue-6.6/series index 238c42ed50..979ae94c2c 100644 --- a/queue-6.6/series +++ b/queue-6.6/series @@ -56,7 +56,6 @@ sunrpc-rpc_clnt_set_transport-must-not-change-the-au.patch sunrpc-rpcbind-should-never-reset-the-port-to-the-va.patch spi-rockchip-fix-register-out-of-bounds-access.patch thermal-drivers-qoriq-power-down-tmu-on-system-suspe.patch -exit-fix-the-usage-of-delay_group_leader-exit_code-i.patch dql-fix-dql-limit-value-when-reset.patch lockdep-fix-wait-context-check-on-softirq-for-preemp.patch objtool-properly-disable-uaccess-validation.patch