October 2024 - Linux-stable-mirror

[PATCH] blk-mq: Make blk_mq_quiesce_tagset() hold the tag list mutex less long

by Bart Van Assche

Make sure that the tag_list_lock mutex is no longer held than necessary. This change reduces latency if e.g. blk_mq_quiesce_tagset() is called concurrently from more than one thread. This function is used by the NVMe core and also by the UFS driver. Reported-by: Peter Wang <peter.wang(a)mediatek.com> Cc: Chao Leng <lengchao(a)huawei.com> Cc: Ming Lei <ming.lei(a)redhat.com> Cc: stable(a)vger.kernel.org Fixes: commit 414dd48e882c ("blk-mq: add tagset quiesce interface") Signed-off-by: Bart Van Assche <bvanassche(a)acm.org> --- block/blk-mq.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/block/blk-mq.c b/block/blk-mq.c index 4b2c8e940f59..1ef227dfb9ba 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -283,8 +283,9 @@ void blk_mq_quiesce_tagset(struct blk_mq_tag_set *set) if (!blk_queue_skip_tagset_quiesce(q)) blk_mq_quiesce_queue_nowait(q); } - blk_mq_wait_quiesce_done(set); mutex_unlock(&set->tag_list_lock); + + blk_mq_wait_quiesce_done(set); } EXPORT_SYMBOL_GPL(blk_mq_quiesce_tagset);

8 months, 4 weeks

4
3
0 0

[PATCH 0/5] cxl: Initialization and shutdown fixes

by Dan Williams

Gregory's modest proposal to fix CXL cxl_mem_probe() failures due to delayed arrival of the CXL "root" infrastructure [1] prompted questions of how the existing mechanism for retrying cxl_mem_probe() could be failing. The critical missing piece in the debug was that Gregory's setup had almost all CXL modules built-in to the kernel. On the way to that discovery several other bugs and init-order corner cases were discovered. The main fix is to make sure the drivers/cxl/Makefile object order supports root CXL ports being fully initialized upon cxl_acpi_probe() exit. The modular case has some similar potential holes that are fixed with MODULE_SOFTDEP() and other fix ups. Finally, an attempt to update cxl_test to reproduce the original report resulted in the discovery of a separate long standing use after free bug in cxl_region_detach(). [1]: http://lore.kernel.org/20241004212504.1246-1-gourry@gourry.net --- Dan Williams (5): cxl/port: Fix CXL port initialization order when the subsystem is built-in cxl/port: Fix cxl_bus_rescan() vs bus_rescan_devices() cxl/acpi: Ensure ports ready at cxl_acpi_probe() return cxl/port: Fix use-after-free, permit out-of-order decoder shutdown cxl/test: Improve init-order fidelity relative to real-world systems drivers/base/core.c | 35 +++++++ drivers/cxl/Kconfig | 1 drivers/cxl/Makefile | 12 +-- drivers/cxl/acpi.c | 7 + drivers/cxl/core/hdm.c | 50 +++++++++-- drivers/cxl/core/port.c | 13 ++- drivers/cxl/core/region.c | 48 +++------- drivers/cxl/cxl.h | 3 - include/linux/device.h | 3 + tools/testing/cxl/test/cxl.c | 200 +++++++++++++++++++++++------------------- tools/testing/cxl/test/mem.c | 1 11 files changed, 228 insertions(+), 145 deletions(-) base-commit: 8cf0b93919e13d1e8d4466eb4080a4c4d9d66d7b

8 months, 4 weeks

5
24
0 0

[merged] nilfs2-fix-kernel-bug-due-to-missing-clearing-of-buffer-delay-flag.patch removed from -mm tree

by Andrew Morton

The quilt patch titled Subject: nilfs2: fix kernel bug due to missing clearing of buffer delay flag has been removed from the -mm tree. Its filename was nilfs2-fix-kernel-bug-due-to-missing-clearing-of-buffer-delay-flag.patch This patch was dropped because it was merged into mainline or a subsystem tree ------------------------------------------------------ From: Ryusuke Konishi <konishi.ryusuke(a)gmail.com> Subject: nilfs2: fix kernel bug due to missing clearing of buffer delay flag Date: Wed, 16 Oct 2024 06:32:07 +0900 Syzbot reported that after nilfs2 reads a corrupted file system image and degrades to read-only, the BUG_ON check for the buffer delay flag in submit_bh_wbc() may fail, causing a kernel bug. This is because the buffer delay flag is not cleared when clearing the buffer state flags to discard a page/folio or a buffer head. So, fix this. This became necessary when the use of nilfs2's own page clear routine was expanded. This state inconsistency does not occur if the buffer is written normally by log writing. Link: https://lkml.kernel.org/r/20241015213300.7114-1-konishi.ryusuke@gmail.com Fixes: 8c26c4e2694a ("nilfs2: fix issue with flush kernel thread after remount in RO mode because of driver's internal error or metadata corruption") Signed-off-by: Ryusuke Konishi <konishi.ryusuke(a)gmail.com> Reported-by: syzbot+985ada84bf055a575c07(a)syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=985ada84bf055a575c07 Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> --- fs/nilfs2/page.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) --- a/fs/nilfs2/page.c~nilfs2-fix-kernel-bug-due-to-missing-clearing-of-buffer-delay-flag +++ a/fs/nilfs2/page.c @@ -77,7 +77,8 @@ void nilfs_forget_buffer(struct buffer_h const unsigned long clear_bits = (BIT(BH_Uptodate) | BIT(BH_Dirty) | BIT(BH_Mapped) | BIT(BH_Async_Write) | BIT(BH_NILFS_Volatile) | - BIT(BH_NILFS_Checked) | BIT(BH_NILFS_Redirected)); + BIT(BH_NILFS_Checked) | BIT(BH_NILFS_Redirected) | + BIT(BH_Delay)); lock_buffer(bh); set_mask_bits(&bh->b_state, clear_bits, 0); @@ -406,7 +407,8 @@ void nilfs_clear_folio_dirty(struct foli const unsigned long clear_bits = (BIT(BH_Uptodate) | BIT(BH_Dirty) | BIT(BH_Mapped) | BIT(BH_Async_Write) | BIT(BH_NILFS_Volatile) | - BIT(BH_NILFS_Checked) | BIT(BH_NILFS_Redirected)); + BIT(BH_NILFS_Checked) | BIT(BH_NILFS_Redirected) | + BIT(BH_Delay)); bh = head; do { _ Patches currently in -mm which might be from konishi.ryusuke(a)gmail.com are nilfs2-fix-kernel-bug-due-to-missing-clearing-of-checked-flag.patch

8 months, 4 weeks

1
0
0 0

[PATCH v2] ASoC: qcom: sc7280: Fix missing Soundwire runtime stream alloc

by Krzysztof Kozlowski

Commit 15c7fab0e047 ("ASoC: qcom: Move Soundwire runtime stream alloc to soundcards") moved the allocation of Soundwire stream runtime from the Qualcomm Soundwire driver to each individual machine sound card driver, except that it forgot to update SC7280 card. Just like for other Qualcomm sound cards using Soundwire, the card driver should allocate and release the runtime. Otherwise sound playback will result in a NULL pointer dereference or other effect of uninitialized memory accesses (which was confirmed on SDM845 having similar issue). Cc: stable(a)vger.kernel.org Cc: Alexey Klimov <alexey.klimov(a)linaro.org> Cc: Steev Klimaszewski <steev(a)kali.org> Fixes: 15c7fab0e047 ("ASoC: qcom: Move Soundwire runtime stream alloc to soundcards") Link: https://lore.kernel.org/r/20241010054109.16938-1-krzysztof.kozlowski@linaro… Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski(a)linaro.org> --- Changes in v2: 1. Add missing 'select SND_SOC_QCOM_SDW' as reported by LKP --- sound/soc/qcom/Kconfig | 1 + sound/soc/qcom/sc7280.c | 10 +++++++++- 2 files changed, 10 insertions(+), 1 deletion(-) diff --git a/sound/soc/qcom/Kconfig b/sound/soc/qcom/Kconfig index 3687b9db5ed4..ca7a30ebd26a 100644 --- a/sound/soc/qcom/Kconfig +++ b/sound/soc/qcom/Kconfig @@ -209,6 +209,7 @@ config SND_SOC_SC7280 tristate "SoC Machine driver for SC7280 boards" depends on I2C && SOUNDWIRE select SND_SOC_QCOM_COMMON + select SND_SOC_QCOM_SDW select SND_SOC_LPASS_SC7280 select SND_SOC_MAX98357A select SND_SOC_WCD938X_SDW diff --git a/sound/soc/qcom/sc7280.c b/sound/soc/qcom/sc7280.c index 207ac5da4dd4..230af8d7b205 100644 --- a/sound/soc/qcom/sc7280.c +++ b/sound/soc/qcom/sc7280.c @@ -23,6 +23,7 @@ #include "common.h" #include "lpass.h" #include "qdsp6/q6afe.h" +#include "sdw.h" #define DEFAULT_MCLK_RATE 19200000 #define RT5682_PLL_FREQ (48000 * 512) @@ -316,6 +317,7 @@ static void sc7280_snd_shutdown(struct snd_pcm_substream *substream) struct snd_soc_card *card = rtd->card; struct sc7280_snd_data *data = snd_soc_card_get_drvdata(card); struct snd_soc_dai *cpu_dai = snd_soc_rtd_to_cpu(rtd, 0); + struct sdw_stream_runtime *sruntime = data->sruntime[cpu_dai->id]; switch (cpu_dai->id) { case MI2S_PRIMARY: @@ -333,6 +335,9 @@ static void sc7280_snd_shutdown(struct snd_pcm_substream *substream) default: break; } + + data->sruntime[cpu_dai->id] = NULL; + sdw_release_stream(sruntime); } static int sc7280_snd_startup(struct snd_pcm_substream *substream) @@ -347,6 +352,8 @@ static int sc7280_snd_startup(struct snd_pcm_substream *substream) switch (cpu_dai->id) { case MI2S_PRIMARY: ret = sc7280_rt5682_init(rtd); + if (ret) + return ret; break; case SECONDARY_MI2S_RX: codec_dai_fmt |= SND_SOC_DAIFMT_NB_NF | SND_SOC_DAIFMT_I2S; @@ -360,7 +367,8 @@ static int sc7280_snd_startup(struct snd_pcm_substream *substream) default: break; } - return ret; + + return qcom_snd_sdw_startup(substream); } static const struct snd_soc_ops sc7280_ops = { -- 2.43.0

8 months, 4 weeks

2
1
0 0

[tip: locking/core] locking/lockdep: Avoid creating new name string literals in lockdep_set_subclass()

by tip-bot2 for Ahmed Ehab

The following commit has been merged into the locking/core branch of tip: Commit-ID: d7fe143cb115076fed0126ad8cf5ba6c3e575e43 Gitweb: https://git.kernel.org/tip/d7fe143cb115076fed0126ad8cf5ba6c3e575e43 Author: Ahmed Ehab <bottaawesome633(a)gmail.com> AuthorDate: Sun, 25 Aug 2024 01:10:30 +03:00 Committer: Boqun Feng <boqun.feng(a)gmail.com> CommitterDate: Thu, 17 Oct 2024 20:07:23 -07:00 locking/lockdep: Avoid creating new name string literals in lockdep_set_subclass() Syzbot reports a problem that a warning will be triggered while searching a lock class in look_up_lock_class(). The cause of the issue is that a new name is created and used by lockdep_set_subclass() instead of using the existing one. This results in a lock instance has a different name pointer than previous registered one stored in lock class, and WARN_ONCE() is triggered because of that in look_up_lock_class(). To fix this, change lockdep_set_subclass() to use the existing name instead of a new one. Hence, no new name will be created by lockdep_set_subclass(). Hence, the warning is avoided. [boqun: Reword the commit log to state the correct issue] Reported-by: <syzbot+7f4a6f7f7051474e40ad(a)syzkaller.appspotmail.com> Fixes: de8f5e4f2dc1f ("lockdep: Introduce wait-type checks") Cc: stable(a)vger.kernel.org Signed-off-by: Ahmed Ehab <bottaawesome633(a)gmail.com> Signed-off-by: Boqun Feng <boqun.feng(a)gmail.com> Link: https://lore.kernel.org/lkml/20240824221031.7751-1-bottaawesome633@gmail.co… --- include/linux/lockdep.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/linux/lockdep.h b/include/linux/lockdep.h index 217f7ab..67964dc 100644 --- a/include/linux/lockdep.h +++ b/include/linux/lockdep.h @@ -173,7 +173,7 @@ static inline void lockdep_init_map(struct lockdep_map *lock, const char *name, (lock)->dep_map.lock_type) #define lockdep_set_subclass(lock, sub) \ - lockdep_init_map_type(&(lock)->dep_map, #lock, (lock)->dep_map.key, sub,\ + lockdep_init_map_type(&(lock)->dep_map, (lock)->dep_map.name, (lock)->dep_map.key, sub,\ (lock)->dep_map.wait_type_inner, \ (lock)->dep_map.wait_type_outer, \ (lock)->dep_map.lock_type)

8 months, 4 weeks

1
0
0 0

[PATCH 5.10] exec: don't WARN for racy path_noexec check

by Thadeu Lima de Souza Cascardo

From: Mateusz Guzik <mjguzik(a)gmail.com> [ Upstream commit 0d196e7589cefe207d5d41f37a0a28a1fdeeb7c6 ] Both i_mode and noexec checks wrapped in WARN_ON stem from an artifact of the previous implementation. They used to legitimately check for the condition, but that got moved up in two commits: 633fb6ac3980 ("exec: move S_ISREG() check earlier") 0fd338b2d2cd ("exec: move path_noexec() check earlier") Instead of being removed said checks are WARN_ON'ed instead, which has some debug value. However, the spurious path_noexec check is racy, resulting in unwarranted warnings should someone race with setting the noexec flag. One can note there is more to perm-checking whether execve is allowed and none of the conditions are guaranteed to still hold after they were tested for. Additionally this does not validate whether the code path did any perm checking to begin with -- it will pass if the inode happens to be regular. Keep the redundant path_noexec() check even though it's mindless nonsense checking for guarantee that isn't given so drop the WARN. Reword the commentary and do small tidy ups while here. Signed-off-by: Mateusz Guzik <mjguzik(a)gmail.com> Link: https://lore.kernel.org/r/20240805131721.765484-1-mjguzik@gmail.com [brauner: keep redundant path_noexec() check] Signed-off-by: Christian Brauner <brauner(a)kernel.org> [cascardo: keep exit label and use it] Signed-off-by: Thadeu Lima de Souza Cascardo <cascardo(a)igalia.com> --- fs/exec.c | 21 +++++++++------------ 1 file changed, 9 insertions(+), 12 deletions(-) diff --git a/fs/exec.c b/fs/exec.c index 6e5324c7e9b6..7144c541818f 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -144,13 +144,11 @@ SYSCALL_DEFINE1(uselib, const char __user *, library) goto out; /* - * may_open() has already checked for this, so it should be - * impossible to trip now. But we need to be extra cautious - * and check again at the very end too. + * Check do_open_execat() for an explanation. */ error = -EACCES; - if (WARN_ON_ONCE(!S_ISREG(file_inode(file)->i_mode) || - path_noexec(&file->f_path))) + if (WARN_ON_ONCE(!S_ISREG(file_inode(file)->i_mode)) || + path_noexec(&file->f_path)) goto exit; fsnotify_open(file); @@ -919,16 +917,16 @@ static struct file *do_open_execat(int fd, struct filename *name, int flags) file = do_filp_open(fd, name, &open_exec_flags); if (IS_ERR(file)) - goto out; + return file; /* - * may_open() has already checked for this, so it should be - * impossible to trip now. But we need to be extra cautious - * and check again at the very end too. + * In the past the regular type check was here. It moved to may_open() in + * 633fb6ac3980 ("exec: move S_ISREG() check earlier"). Since then it is + * an invariant that all non-regular files error out before we get here. */ err = -EACCES; - if (WARN_ON_ONCE(!S_ISREG(file_inode(file)->i_mode) || - path_noexec(&file->f_path))) + if (WARN_ON_ONCE(!S_ISREG(file_inode(file)->i_mode)) || + path_noexec(&file->f_path)) goto exit; err = deny_write_access(file); @@ -938,7 +936,6 @@ static struct file *do_open_execat(int fd, struct filename *name, int flags) if (name->name[0] != '\0') fsnotify_open(file); -out: return file; exit: -- 2.34.1

8 months, 4 weeks

1
0
0 0

[PATCH 5.15] exec: don't WARN for racy path_noexec check

by Thadeu Lima de Souza Cascardo

From: Mateusz Guzik <mjguzik(a)gmail.com> [ Upstream commit 0d196e7589cefe207d5d41f37a0a28a1fdeeb7c6 ] Both i_mode and noexec checks wrapped in WARN_ON stem from an artifact of the previous implementation. They used to legitimately check for the condition, but that got moved up in two commits: 633fb6ac3980 ("exec: move S_ISREG() check earlier") 0fd338b2d2cd ("exec: move path_noexec() check earlier") Instead of being removed said checks are WARN_ON'ed instead, which has some debug value. However, the spurious path_noexec check is racy, resulting in unwarranted warnings should someone race with setting the noexec flag. One can note there is more to perm-checking whether execve is allowed and none of the conditions are guaranteed to still hold after they were tested for. Additionally this does not validate whether the code path did any perm checking to begin with -- it will pass if the inode happens to be regular. Keep the redundant path_noexec() check even though it's mindless nonsense checking for guarantee that isn't given so drop the WARN. Reword the commentary and do small tidy ups while here. Signed-off-by: Mateusz Guzik <mjguzik(a)gmail.com> Link: https://lore.kernel.org/r/20240805131721.765484-1-mjguzik@gmail.com [brauner: keep redundant path_noexec() check] Signed-off-by: Christian Brauner <brauner(a)kernel.org> [cascardo: keep exit label and use it] Signed-off-by: Thadeu Lima de Souza Cascardo <cascardo(a)igalia.com> --- fs/exec.c | 21 +++++++++------------ 1 file changed, 9 insertions(+), 12 deletions(-) diff --git a/fs/exec.c b/fs/exec.c index 26f0b79cb4f9..8395e7ff7b94 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -142,13 +142,11 @@ SYSCALL_DEFINE1(uselib, const char __user *, library) goto out; /* - * may_open() has already checked for this, so it should be - * impossible to trip now. But we need to be extra cautious - * and check again at the very end too. + * Check do_open_execat() for an explanation. */ error = -EACCES; - if (WARN_ON_ONCE(!S_ISREG(file_inode(file)->i_mode) || - path_noexec(&file->f_path))) + if (WARN_ON_ONCE(!S_ISREG(file_inode(file)->i_mode)) || + path_noexec(&file->f_path)) goto exit; fsnotify_open(file); @@ -919,16 +917,16 @@ static struct file *do_open_execat(int fd, struct filename *name, int flags) file = do_filp_open(fd, name, &open_exec_flags); if (IS_ERR(file)) - goto out; + return file; /* - * may_open() has already checked for this, so it should be - * impossible to trip now. But we need to be extra cautious - * and check again at the very end too. + * In the past the regular type check was here. It moved to may_open() in + * 633fb6ac3980 ("exec: move S_ISREG() check earlier"). Since then it is + * an invariant that all non-regular files error out before we get here. */ err = -EACCES; - if (WARN_ON_ONCE(!S_ISREG(file_inode(file)->i_mode) || - path_noexec(&file->f_path))) + if (WARN_ON_ONCE(!S_ISREG(file_inode(file)->i_mode)) || + path_noexec(&file->f_path)) goto exit; err = deny_write_access(file); @@ -938,7 +936,6 @@ static struct file *do_open_execat(int fd, struct filename *name, int flags) if (name->name[0] != '\0') fsnotify_open(file); -out: return file; exit: -- 2.34.1

8 months, 4 weeks

1
0
0 0

[PATCH 6.1] exec: don't WARN for racy path_noexec check

by Thadeu Lima de Souza Cascardo

From: Mateusz Guzik <mjguzik(a)gmail.com> [ Upstream commit 0d196e7589cefe207d5d41f37a0a28a1fdeeb7c6 ] Both i_mode and noexec checks wrapped in WARN_ON stem from an artifact of the previous implementation. They used to legitimately check for the condition, but that got moved up in two commits: 633fb6ac3980 ("exec: move S_ISREG() check earlier") 0fd338b2d2cd ("exec: move path_noexec() check earlier") Instead of being removed said checks are WARN_ON'ed instead, which has some debug value. However, the spurious path_noexec check is racy, resulting in unwarranted warnings should someone race with setting the noexec flag. One can note there is more to perm-checking whether execve is allowed and none of the conditions are guaranteed to still hold after they were tested for. Additionally this does not validate whether the code path did any perm checking to begin with -- it will pass if the inode happens to be regular. Keep the redundant path_noexec() check even though it's mindless nonsense checking for guarantee that isn't given so drop the WARN. Reword the commentary and do small tidy ups while here. Signed-off-by: Mateusz Guzik <mjguzik(a)gmail.com> Link: https://lore.kernel.org/r/20240805131721.765484-1-mjguzik@gmail.com [brauner: keep redundant path_noexec() check] Signed-off-by: Christian Brauner <brauner(a)kernel.org> [cascardo: keep exit label and use it] Signed-off-by: Thadeu Lima de Souza Cascardo <cascardo(a)igalia.com> --- fs/exec.c | 21 +++++++++------------ 1 file changed, 9 insertions(+), 12 deletions(-) diff --git a/fs/exec.c b/fs/exec.c index 65d3ebc24fd3..a42c9b8b070d 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -141,13 +141,11 @@ SYSCALL_DEFINE1(uselib, const char __user *, library) goto out; /* - * may_open() has already checked for this, so it should be - * impossible to trip now. But we need to be extra cautious - * and check again at the very end too. + * Check do_open_execat() for an explanation. */ error = -EACCES; - if (WARN_ON_ONCE(!S_ISREG(file_inode(file)->i_mode) || - path_noexec(&file->f_path))) + if (WARN_ON_ONCE(!S_ISREG(file_inode(file)->i_mode)) || + path_noexec(&file->f_path)) goto exit; fsnotify_open(file); @@ -927,16 +925,16 @@ static struct file *do_open_execat(int fd, struct filename *name, int flags) file = do_filp_open(fd, name, &open_exec_flags); if (IS_ERR(file)) - goto out; + return file; /* - * may_open() has already checked for this, so it should be - * impossible to trip now. But we need to be extra cautious - * and check again at the very end too. + * In the past the regular type check was here. It moved to may_open() in + * 633fb6ac3980 ("exec: move S_ISREG() check earlier"). Since then it is + * an invariant that all non-regular files error out before we get here. */ err = -EACCES; - if (WARN_ON_ONCE(!S_ISREG(file_inode(file)->i_mode) || - path_noexec(&file->f_path))) + if (WARN_ON_ONCE(!S_ISREG(file_inode(file)->i_mode)) || + path_noexec(&file->f_path)) goto exit; err = deny_write_access(file); @@ -946,7 +944,6 @@ static struct file *do_open_execat(int fd, struct filename *name, int flags) if (name->name[0] != '\0') fsnotify_open(file); -out: return file; exit: -- 2.34.1

8 months, 4 weeks

1
0
0 0

[PATCH 6.6] exec: don't WARN for racy path_noexec check

by Thadeu Lima de Souza Cascardo

From: Mateusz Guzik <mjguzik(a)gmail.com> [ Upstream commit 0d196e7589cefe207d5d41f37a0a28a1fdeeb7c6 ] Both i_mode and noexec checks wrapped in WARN_ON stem from an artifact of the previous implementation. They used to legitimately check for the condition, but that got moved up in two commits: 633fb6ac3980 ("exec: move S_ISREG() check earlier") 0fd338b2d2cd ("exec: move path_noexec() check earlier") Instead of being removed said checks are WARN_ON'ed instead, which has some debug value. However, the spurious path_noexec check is racy, resulting in unwarranted warnings should someone race with setting the noexec flag. One can note there is more to perm-checking whether execve is allowed and none of the conditions are guaranteed to still hold after they were tested for. Additionally this does not validate whether the code path did any perm checking to begin with -- it will pass if the inode happens to be regular. Keep the redundant path_noexec() check even though it's mindless nonsense checking for guarantee that isn't given so drop the WARN. Reword the commentary and do small tidy ups while here. Signed-off-by: Mateusz Guzik <mjguzik(a)gmail.com> Link: https://lore.kernel.org/r/20240805131721.765484-1-mjguzik@gmail.com [brauner: keep redundant path_noexec() check] Signed-off-by: Christian Brauner <brauner(a)kernel.org> [cascardo: keep exit label and use it] Signed-off-by: Thadeu Lima de Souza Cascardo <cascardo(a)igalia.com> --- fs/exec.c | 21 +++++++++------------ 1 file changed, 9 insertions(+), 12 deletions(-) diff --git a/fs/exec.c b/fs/exec.c index f49b352a6032..7776209d98c1 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -143,13 +143,11 @@ SYSCALL_DEFINE1(uselib, const char __user *, library) goto out; /* - * may_open() has already checked for this, so it should be - * impossible to trip now. But we need to be extra cautious - * and check again at the very end too. + * Check do_open_execat() for an explanation. */ error = -EACCES; - if (WARN_ON_ONCE(!S_ISREG(file_inode(file)->i_mode) || - path_noexec(&file->f_path))) + if (WARN_ON_ONCE(!S_ISREG(file_inode(file)->i_mode)) || + path_noexec(&file->f_path)) goto exit; error = -ENOEXEC; @@ -925,23 +923,22 @@ static struct file *do_open_execat(int fd, struct filename *name, int flags) file = do_filp_open(fd, name, &open_exec_flags); if (IS_ERR(file)) - goto out; + return file; /* - * may_open() has already checked for this, so it should be - * impossible to trip now. But we need to be extra cautious - * and check again at the very end too. + * In the past the regular type check was here. It moved to may_open() in + * 633fb6ac3980 ("exec: move S_ISREG() check earlier"). Since then it is + * an invariant that all non-regular files error out before we get here. */ err = -EACCES; - if (WARN_ON_ONCE(!S_ISREG(file_inode(file)->i_mode) || - path_noexec(&file->f_path))) + if (WARN_ON_ONCE(!S_ISREG(file_inode(file)->i_mode)) || + path_noexec(&file->f_path)) goto exit; err = deny_write_access(file); if (err) goto exit; -out: return file; exit: -- 2.34.1

8 months, 4 weeks

1
0
0 0

[PATCH 5.15 00/82] 5.15.169-rc1 review

by Greg Kroah-Hartman

This is the start of the stable review cycle for the 5.15.169 release. There are 82 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know. Responses should be made by Wed, 23 Oct 2024 10:22:25 +0000. Anything received after that time might be too late. The whole patch series can be found in one patch at: https://www.kernel.org/pub/linux/kernel/v5.x/stable-review/patch-5.15.169-r… or in the git tree and branch at: git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git linux-5.15.y and the diffstat can be found below. thanks, greg k-h ------------- Pseudo-Shortlog of commits: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Linux 5.15.169-rc1 Vasiliy Kovalev <kovalev(a)altlinux.org> ALSA: hda/conexant - Use cached pin control for Node 0x1d on HP EliteOne 1000 G2 Aneesh Kumar K.V <aneesh.kumar(a)linux.ibm.com> powerpc/mm: Always update max/min_low_pfn in mem_topology_setup() Ryusuke Konishi <konishi.ryusuke(a)gmail.com> nilfs2: propagate directory read errors from nilfs_find_entry() Paolo Abeni <pabeni(a)redhat.com> mptcp: prevent MPC handshake on port-based signal endpoints Matthieu Baerts (NGI0) <matttbe(a)kernel.org> mptcp: pm: fix UaF read in mptcp_pm_nl_rm_addr_or_subflow Matthieu Baerts (NGI0) <matttbe(a)kernel.org> mptcp: fallback when MPTCP opts are dropped after 1st data Paolo Abeni <pabeni(a)redhat.com> tcp: fix mptcp DSS corruption due to large pmtu xmit Paolo Abeni <pabeni(a)redhat.com> mptcp: handle consistently DSS corruption Geliang Tang <geliang.tang(a)suse.com> mptcp: track and update contiguous data status Marc Zyngier <maz(a)kernel.org> irqchip/gic-v4: Don't allow a VMOVP on a dying VPE Sergey Matsievskiy <matsievskiysv(a)gmail.com> pinctrl: ocelot: fix system hang on level based interrupts Pawan Gupta <pawan.kumar.gupta(a)linux.intel.com> x86/entry_32: Clear CPU buffers after register restore in NMI return Pawan Gupta <pawan.kumar.gupta(a)linux.intel.com> x86/entry_32: Do not clobber user EFLAGS.ZF Zhang Rui <rui.zhang(a)intel.com> x86/apic: Always explicitly disarm TSC-deadline timer Nathan Chancellor <nathan(a)kernel.org> x86/resctrl: Annotate get_mem_config() functions as __init Takashi Iwai <tiwai(a)suse.de> parport: Proper fix for array out-of-bounds access Daniele Palmas <dnlplm(a)gmail.com> USB: serial: option: add Telit FN920C04 MBIM compositions Benjamin B. Frost <benjamin(a)geanix.com> USB: serial: option: add support for Quectel EG916Q-GL Mathias Nyman <mathias.nyman(a)linux.intel.com> xhci: Mitigate failed set dequeue pointer commands Mathias Nyman <mathias.nyman(a)linux.intel.com> xhci: Fix incorrect stream context type macro Luiz Augusto von Dentz <luiz.von.dentz(a)intel.com> Bluetooth: btusb: Fix regression with fake CSR controllers 0a12:0001 Aaron Thompson <dev(a)aaront.org> Bluetooth: Remove debugfs directory on module init failure Javier Carrasco <javier.carrasco.cruz(a)gmail.com> iio: adc: ti-ads124s08: add missing select IIO_(TRIGGERED_)BUFFER in Kconfig Javier Carrasco <javier.carrasco.cruz(a)gmail.com> iio: proximity: mb1232: add missing select IIO_(TRIGGERED_)BUFFER in Kconfig Emil Gedenryd <emil.gedenryd(a)axis.com> iio: light: opt3001: add missing full-scale range value Javier Carrasco <javier.carrasco.cruz(a)gmail.com> iio: light: veml6030: fix IIO device retrieval from embedded device Javier Carrasco <javier.carrasco.cruz(a)gmail.com> iio: light: veml6030: fix ALS sensor resolution Christophe JAILLET <christophe.jaillet(a)wanadoo.fr> iio: hid-sensors: Fix an error handling path in _hid_sensor_set_report_latency() Javier Carrasco <javier.carrasco.cruz(a)gmail.com> iio: adc: ti-ads8688: add missing select IIO_(TRIGGERED_)BUFFER in Kconfig Javier Carrasco <javier.carrasco.cruz(a)gmail.com> iio: dac: stm32-dac-core: add missing select REGMAP_MMIO in Kconfig Javier Carrasco <javier.carrasco.cruz(a)gmail.com> iio: dac: ltc1660: add missing select REGMAP_SPI in Kconfig Javier Carrasco <javier.carrasco.cruz(a)gmail.com> iio: dac: ad5770r: add missing select REGMAP_SPI in Kconfig Nikolay Kuratov <kniv(a)yandex-team.ru> drm/vmwgfx: Handle surface check failure correctly Ville Syrjälä <ville.syrjala(a)linux.intel.com> drm/radeon: Fix encoder->possible_clones Jens Axboe <axboe(a)kernel.dk> io_uring/sqpoll: close race on waiting for sqring entries Omar Sandoval <osandov(a)fb.com> blk-rq-qos: fix crash on rq_qos_wait vs. rq_qos_wake_function race Johannes Wikner <kwikner(a)ethz.ch> x86/bugs: Do not use UNTRAIN_RET with IBPB on entry Johannes Wikner <kwikner(a)ethz.ch> x86/bugs: Skip RSB fill at VMEXIT Johannes Wikner <kwikner(a)ethz.ch> x86/entry: Have entry_ibpb() invalidate return predictions Johannes Wikner <kwikner(a)ethz.ch> x86/cpufeatures: Add a IBPB_NO_RET BUG flag Jim Mattson <jmattson(a)google.com> x86/cpufeatures: Define X86_FEATURE_AMD_IBPB_RET Michael Mueller <mimu(a)linux.ibm.com> KVM: s390: Change virtual to physical address access in diag 0x258 handler Thomas Weißschuh <thomas.weissschuh(a)linutronix.de> s390/sclp_vt220: Convert newlines to CRLF instead of LFCR Lu Baolu <baolu.lu(a)linux.intel.com> iommu/vt-d: Fix incorrect pci_for_each_dma_alias() for non-PCI devices Felix Moessbauer <felix.moessbauer(a)siemens.com> io_uring/sqpoll: do not put cpumask on stack Jens Axboe <axboe(a)kernel.dk> io_uring/sqpoll: retain test for whether the CPU is valid Felix Moessbauer <felix.moessbauer(a)siemens.com> io_uring/sqpoll: do not allow pinning outside of cpuset Wachowski, Karol <karol.wachowski(a)intel.com> drm/shmem-helper: Fix BUG_ON() on mmap(PROT_WRITE, MAP_PRIVATE) Breno Leitao <leitao(a)debian.org> KVM: Fix a data race on last_boosted_vcpu in kvm_vcpu_on_spin() Mikulas Patocka <mpatocka(a)redhat.com> dm-crypt, dm-verity: disable tasklets Johannes Berg <johannes.berg(a)intel.com> wifi: mac80211: fix potential key use-after-free Patrick Roy <roypat(a)amazon.co.uk> secretmem: disable memfd_secret() if arch cannot set direct map Liu Shixin <liushixin2(a)huawei.com> mm/swapfile: skip HugeTLB pages for unuse_vma OGAWA Hirofumi <hirofumi(a)mail.parknet.co.jp> fat: fix uninitialized variable Nianyao Tang <tangnianyao(a)huawei.com> irqchip/gic-v3-its: Fix VSYNC referencing an unmapped VPE on GIC v4.1 Oleksij Rempel <linux(a)rempel-privat.de> net: macb: Avoid 20s boot delay by skipping MDIO bus registration for fixed-link PHY Mark Rutland <mark.rutland(a)arm.com> arm64: probes: Fix simulate_ldr*_literal() Mark Rutland <mark.rutland(a)arm.com> arm64: probes: Remove broken LDR (literal) uprobe support Jinjie Ruan <ruanjinjie(a)huawei.com> posix-clock: Fix missing timespec64 check in pc_clock_settime() Wei Fang <wei.fang(a)nxp.com> net: enetc: add missing static descriptor and inline keyword Wei Fang <wei.fang(a)nxp.com> net: enetc: remove xdp_drops statistic from enetc_xdp_drop() Jan Kara <jack(a)suse.cz> udf: Fix bogus checksum computation in udf_rename() Jan Kara <jack(a)suse.cz> udf: Don't return bh from udf_expand_dir_adinicb() Jan Kara <jack(a)suse.cz> udf: Handle error when expanding directory Jan Kara <jack(a)suse.cz> udf: Remove old directory iteration code Jan Kara <jack(a)suse.cz> udf: Convert udf_link() to new directory iteration code Jan Kara <jack(a)suse.cz> udf: Convert udf_mkdir() to new directory iteration code Jan Kara <jack(a)suse.cz> udf: Convert udf_add_nondir() to new directory iteration Jan Kara <jack(a)suse.cz> udf: Implement adding of dir entries using new iteration code Jan Kara <jack(a)suse.cz> udf: Convert udf_unlink() to new directory iteration code Jan Kara <jack(a)suse.cz> udf: Convert udf_rmdir() to new directory iteration code Jan Kara <jack(a)suse.cz> udf: Convert empty_dir() to new directory iteration code Jan Kara <jack(a)suse.cz> udf: Convert udf_get_parent() to new directory iteration code Jan Kara <jack(a)suse.cz> udf: Convert udf_lookup() to use new directory iteration code Jan Kara <jack(a)suse.cz> udf: Convert udf_readdir() to new directory iteration Jan Kara <jack(a)suse.cz> udf: Convert udf_rename() to new directory iteration code Jan Kara <jack(a)suse.cz> udf: Provide function to mark entry as deleted using new directory iteration code Jan Kara <jack(a)suse.cz> udf: Implement searching for directory entry using new iteration code Jan Kara <jack(a)suse.cz> udf: Move udf_expand_dir_adinicb() to its callsite Jan Kara <jack(a)suse.cz> udf: Convert udf_expand_dir_adinicb() to new directory iteration Jan Kara <jack(a)suse.cz> udf: New directory iteration code Vasiliy Kovalev <kovalev(a)altlinux.org> ALSA: hda/conexant - Fix audio routing for HP EliteOne 1000 G2 ------------- Diffstat: Makefile | 4 +- arch/arm64/kernel/probes/decode-insn.c | 16 +- arch/arm64/kernel/probes/simulate-insn.c | 18 +- arch/powerpc/mm/numa.c | 6 +- arch/s390/kvm/diag.c | 2 +- arch/x86/entry/entry.S | 5 + arch/x86/entry/entry_32.S | 6 +- arch/x86/include/asm/cpufeatures.h | 4 +- arch/x86/kernel/apic/apic.c | 14 +- arch/x86/kernel/cpu/bugs.c | 32 + arch/x86/kernel/cpu/common.c | 3 + arch/x86/kernel/cpu/resctrl/core.c | 4 +- block/blk-rq-qos.c | 2 +- drivers/bluetooth/btusb.c | 13 +- drivers/gpu/drm/drm_gem_shmem_helper.c | 3 + drivers/gpu/drm/radeon/radeon_encoders.c | 2 +- drivers/gpu/drm/vmwgfx/vmwgfx_kms.c | 1 + drivers/iio/adc/Kconfig | 4 + .../iio/common/hid-sensors/hid-sensor-trigger.c | 2 +- drivers/iio/dac/Kconfig | 3 + drivers/iio/light/opt3001.c | 4 + drivers/iio/light/veml6030.c | 5 +- drivers/iio/proximity/Kconfig | 2 + drivers/iommu/intel/iommu.c | 4 +- drivers/irqchip/irq-gic-v3-its.c | 26 +- drivers/md/dm-crypt.c | 37 +- drivers/net/ethernet/cadence/macb_main.c | 14 +- drivers/net/ethernet/freescale/enetc/enetc.c | 2 +- drivers/parport/procfs.c | 22 +- drivers/pinctrl/pinctrl-ocelot.c | 8 +- drivers/s390/char/sclp_vt220.c | 4 +- drivers/usb/host/xhci-ring.c | 2 +- drivers/usb/host/xhci.h | 2 +- drivers/usb/serial/option.c | 8 + fs/fat/namei_vfat.c | 2 +- fs/nilfs2/dir.c | 50 +- fs/nilfs2/namei.c | 39 +- fs/nilfs2/nilfs.h | 2 +- fs/udf/dir.c | 148 +-- fs/udf/directory.c | 594 ++++++++--- fs/udf/inode.c | 90 -- fs/udf/namei.c | 1038 +++++++------------- fs/udf/udfdecl.h | 45 +- include/linux/fsl/enetc_mdio.h | 3 +- include/linux/irqchip/arm-gic-v4.h | 4 +- io_uring/io_uring.c | 21 +- kernel/time/posix-clock.c | 3 + mm/secretmem.c | 4 +- mm/swapfile.c | 2 +- net/bluetooth/af_bluetooth.c | 1 + net/ipv4/tcp_output.c | 2 +- net/mac80211/cfg.c | 3 + net/mac80211/key.c | 2 +- net/mptcp/mib.c | 3 + net/mptcp/mib.h | 3 + net/mptcp/pm_netlink.c | 3 +- net/mptcp/protocol.c | 23 +- net/mptcp/protocol.h | 2 + net/mptcp/subflow.c | 19 +- sound/pci/hda/patch_conexant.c | 19 + virt/kvm/kvm_main.c | 5 +- 61 files changed, 1172 insertions(+), 1242 deletions(-)

8 months, 4 weeks

8
89
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror October 2024