March 2023 - Linux-stable-mirror

[PATCH] riscv: Handle zicsr/zifencei issues between clang and binutils

by Nathan Chancellor

There are two related issues that appear in certain combinations with clang and GNU binutils. The first occurs when a version of clang that supports zicsr or zifencei via '-march=' [1] (i.e, >= 17.x) is used in combination with a version of GNU binutils that do not recognize zicsr and zifencei in the '-march=' value (i.e., < 2.36): riscv64-linux-gnu-ld: -march=rv64i2p0_m2p0_a2p0_c2p0_zicsr2p0_zifencei2p0: Invalid or unknown z ISA extension: 'zifencei' riscv64-linux-gnu-ld: failed to merge target specific data of file fs/efivarfs/file.o riscv64-linux-gnu-ld: -march=rv64i2p0_m2p0_a2p0_c2p0_zicsr2p0_zifencei2p0: Invalid or unknown z ISA extension: 'zifencei' riscv64-linux-gnu-ld: failed to merge target specific data of file fs/efivarfs/super.o The second occurs when a version of clang that does not support zicsr or zifencei via '-march=' (i.e., <= 16.x) is used in combination with a version of GNU as that defaults to a newer ISA base spec, which requires specifying zicsr and zifencei in the '-march=' value explicitly (i.e, >= 2.38): ../arch/riscv/kernel/kexec_relocate.S: Assembler messages: ../arch/riscv/kernel/kexec_relocate.S:147: Error: unrecognized opcode `fence.i', extension `zifencei' required clang-12: error: assembler command failed with exit code 1 (use -v to see invocation) This is the same issue addressed by commit 6df2a016c0c8 ("riscv: fix build with binutils 2.38") (see [2] for additional information) but older versions of clang miss out on it because the cc-option check fails: clang-12: error: invalid arch name 'rv64imac_zicsr_zifencei', unsupported standard user-level extension 'zicsr' clang-12: error: invalid arch name 'rv64imac_zicsr_zifencei', unsupported standard user-level extension 'zicsr' To resolve the first issue, only attempt to add zicsr and zifencei to the march string when using the GNU assembler 2.38 or newer, which is when the default ISA spec was updated, requiring these extensions to be specified explicitly. LLVM implements an older version of the base specification for all currently released versions, so these instructions are available as part of the 'i' extension. If LLVM's implementation is updated in the future, a CONFIG_AS_IS_LLVM condition can be added to CONFIG_TOOLCHAIN_NEEDS_EXPLICIT_ZICSR_ZIFENCEI. To resolve the second issue, use version 2.2 of the base ISA spec when using an older version of clang that does not support zicsr or zifencei via '-march=', as that is the spec version most compatible with the one clang/LLVM implements and avoids the need to specify zicsr and zifencei explicitly due to still being a part of 'i'. [1]: https://github.com/llvm/llvm-project/commit/22e199e6afb1263c943c0c0d4498694… [2]: https://lore.kernel.org/ZAxT7T9Xy1Fo3d5W@aurel32.net/ Cc: stable(a)vger.kernel.org Link: https://github.com/ClangBuiltLinux/linux/issues/1808 Co-developed-by: Conor Dooley <conor.dooley(a)microchip.com> Signed-off-by: Conor Dooley <conor.dooley(a)microchip.com> Signed-off-by: Nathan Chancellor <nathan(a)kernel.org> --- This is essentially a v3 of Conor's v1 and v2 but since I am sending the patch after finding a separate but related issue, I left it at v1: - v1: https://lore.kernel.org/20230223220546.52879-1-conor@kernel.org/ - v2: https://lore.kernel.org/20230308220842.1231003-1-conor@kernel.org/ I have built allmodconfig with the following toolchain combinations to confirm this problem is resolved: - clang 12/17 + GNU as and ld 2.35/2.39 - clang 12/17 with the integrated assembler + GNU ld 2.35/2.39 - clang 12/17 with the integrated assembler + ld.lld There are a couple of other incompatibilities between clang-17 and GNU binutils that I had to patch to get allmodconfig to build successfully but those are less likely to be hit in practice because the full LLVM stack can be used with LLVM versions 13.x and newer. I will follow up with separate issues and patches. --- arch/riscv/Kconfig | 22 ++++++++++++++++++++++ arch/riscv/Makefile | 10 ++++++---- 2 files changed, 28 insertions(+), 4 deletions(-) diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig index c5e42cc37604..5b182d1c196c 100644 --- a/arch/riscv/Kconfig +++ b/arch/riscv/Kconfig @@ -464,6 +464,28 @@ config TOOLCHAIN_HAS_ZIHINTPAUSE depends on !32BIT || $(cc-option,-mabi=ilp32 -march=rv32ima_zihintpause) depends on LLD_VERSION >= 150000 || LD_VERSION >= 23600 +config TOOLCHAIN_NEEDS_EXPLICIT_ZICSR_ZIFENCEI + def_bool y + # https://sourceware.org/git/?p=binutils-gdb.git;a=commit;h=aed44286efa8ae871… + depends on AS_IS_GNU && AS_VERSION >= 23800 + help + Newer binutils versions default to ISA spec version 20191213 which + moves some instructions from the I extension to the Zicsr and Zifencei + extensions. + +config TOOLCHAIN_NEEDS_OLD_ISA_SPEC + def_bool y + depends on TOOLCHAIN_NEEDS_EXPLICIT_ZICSR_ZIFENCEI + # https://github.com/llvm/llvm-project/commit/22e199e6afb1263c943c0c0d4498694… + depends on CC_IS_CLANG && CLANG_VERSION < 170000 + help + Certain versions of clang do not support zicsr and zifencei via -march + but newer versions of binutils require it for the reasons noted in the + help text of CONFIG_TOOLCHAIN_NEEDS_EXPLICIT_ZICSR_ZIFENCEI. This + option causes an older ISA spec compatible with these older versions + of clang to be passed to GAS, which has the same result as passing zicsr + and zifencei to -march. + config FPU bool "FPU support" default y diff --git a/arch/riscv/Makefile b/arch/riscv/Makefile index 4de83b9b1772..b05e833a022d 100644 --- a/arch/riscv/Makefile +++ b/arch/riscv/Makefile @@ -57,10 +57,12 @@ riscv-march-$(CONFIG_ARCH_RV64I) := rv64ima riscv-march-$(CONFIG_FPU) := $(riscv-march-y)fd riscv-march-$(CONFIG_RISCV_ISA_C) := $(riscv-march-y)c -# Newer binutils versions default to ISA spec version 20191213 which moves some -# instructions from the I extension to the Zicsr and Zifencei extensions. -toolchain-need-zicsr-zifencei := $(call cc-option-yn, -march=$(riscv-march-y)_zicsr_zifencei) -riscv-march-$(toolchain-need-zicsr-zifencei) := $(riscv-march-y)_zicsr_zifencei +ifdef CONFIG_TOOLCHAIN_NEEDS_OLD_ISA_SPEC +KBUILD_CFLAGS += -Wa,-misa-spec=2.2 +KBUILD_AFLAGS += -Wa,-misa-spec=2.2 +else +riscv-march-$(CONFIG_TOOLCHAIN_NEEDS_EXPLICIT_ZICSR_ZIFENCEI) := $(riscv-march-y)_zicsr_zifencei +endif # Check if the toolchain supports Zihintpause extension riscv-march-$(CONFIG_TOOLCHAIN_HAS_ZIHINTPAUSE) := $(riscv-march-y)_zihintpause --- base-commit: eeac8ede17557680855031c6f305ece2378af326 change-id: 20230313-riscv-zicsr-zifencei-fiasco-2941caebe7dc Best regards, -- Nathan Chancellor <nathan(a)kernel.org>

1 year, 9 months

5
5
0 0

patch "usb: chipidea: core: fix possible concurrent when switch role" added to usb-linus

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled usb: chipidea: core: fix possible concurrent when switch role to my usb git tree which can be found at git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb.git in the usb-linus branch. The patch will show up in the next release of the linux-next tree (usually sometime within the next 24 hours during the week.) The patch will hopefully also be merged in Linus's tree for the next -rc kernel release. If you have any questions about this process, please let me know. From 451b15ed138ec15bffbebb58a00ebdd884c3e659 Mon Sep 17 00:00:00 2001 From: Xu Yang <xu.yang_2(a)nxp.com> Date: Fri, 17 Mar 2023 14:15:16 +0800 Subject: usb: chipidea: core: fix possible concurrent when switch role The user may call role_store() when driver is handling ci_handle_id_switch() which is triggerred by otg event or power lost event. Unfortunately, the controller may go into chaos in this case. Fix this by protecting it with mutex lock. Fixes: a932a8041ff9 ("usb: chipidea: core: add sysfs group") cc: <stable(a)vger.kernel.org> Acked-by: Peter Chen <peter.chen(a)kernel.org> Signed-off-by: Xu Yang <xu.yang_2(a)nxp.com> Link: https://lore.kernel.org/r/20230317061516.2451728-2-xu.yang_2@nxp.com Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/usb/chipidea/ci.h | 2 ++ drivers/usb/chipidea/core.c | 8 +++++++- drivers/usb/chipidea/otg.c | 5 ++++- 3 files changed, 13 insertions(+), 2 deletions(-) diff --git a/drivers/usb/chipidea/ci.h b/drivers/usb/chipidea/ci.h index 005c67cb3afb..f210b7489fd5 100644 --- a/drivers/usb/chipidea/ci.h +++ b/drivers/usb/chipidea/ci.h @@ -208,6 +208,7 @@ struct hw_bank { * @in_lpm: if the core in low power mode * @wakeup_int: if wakeup interrupt occur * @rev: The revision number for controller + * @mutex: protect code from concorrent running when doing role switch */ struct ci_hdrc { struct device *dev; @@ -260,6 +261,7 @@ struct ci_hdrc { bool in_lpm; bool wakeup_int; enum ci_revision rev; + struct mutex mutex; }; static inline struct ci_role_driver *ci_role(struct ci_hdrc *ci) diff --git a/drivers/usb/chipidea/core.c b/drivers/usb/chipidea/core.c index b6f2a41de20e..281fc51720ce 100644 --- a/drivers/usb/chipidea/core.c +++ b/drivers/usb/chipidea/core.c @@ -987,8 +987,12 @@ static ssize_t role_store(struct device *dev, if (role == CI_ROLE_END) return -EINVAL; - if (role == ci->role) + mutex_lock(&ci->mutex); + + if (role == ci->role) { + mutex_unlock(&ci->mutex); return n; + } pm_runtime_get_sync(dev); disable_irq(ci->irq); @@ -998,6 +1002,7 @@ static ssize_t role_store(struct device *dev, ci_handle_vbus_change(ci); enable_irq(ci->irq); pm_runtime_put_sync(dev); + mutex_unlock(&ci->mutex); return (ret == 0) ? n : ret; } @@ -1033,6 +1038,7 @@ static int ci_hdrc_probe(struct platform_device *pdev) return -ENOMEM; spin_lock_init(&ci->lock); + mutex_init(&ci->mutex); ci->dev = dev; ci->platdata = dev_get_platdata(dev); ci->imx28_write_fix = !!(ci->platdata->flags & diff --git a/drivers/usb/chipidea/otg.c b/drivers/usb/chipidea/otg.c index 622c3b68aa1e..f5490f2a5b6b 100644 --- a/drivers/usb/chipidea/otg.c +++ b/drivers/usb/chipidea/otg.c @@ -167,8 +167,10 @@ static int hw_wait_vbus_lower_bsv(struct ci_hdrc *ci) void ci_handle_id_switch(struct ci_hdrc *ci) { - enum ci_role role = ci_otg_role(ci); + enum ci_role role; + mutex_lock(&ci->mutex); + role = ci_otg_role(ci); if (role != ci->role) { dev_dbg(ci->dev, "switching from %s to %s\n", ci_role(ci)->name, ci->roles[role]->name); @@ -198,6 +200,7 @@ void ci_handle_id_switch(struct ci_hdrc *ci) if (role == CI_ROLE_GADGET) ci_handle_vbus_change(ci); } + mutex_unlock(&ci->mutex); } /** * ci_otg_work - perform otg (vbus/id) event handle -- 2.40.0

1 year, 9 months

1
0
0 0

patch "usb: chipdea: core: fix return -EINVAL if request role is the same" added to usb-linus

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled usb: chipdea: core: fix return -EINVAL if request role is the same to my usb git tree which can be found at git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb.git in the usb-linus branch. The patch will show up in the next release of the linux-next tree (usually sometime within the next 24 hours during the week.) The patch will hopefully also be merged in Linus's tree for the next -rc kernel release. If you have any questions about this process, please let me know. From 3670de80678961eda7fa2220883fc77c16868951 Mon Sep 17 00:00:00 2001 From: Xu Yang <xu.yang_2(a)nxp.com> Date: Fri, 17 Mar 2023 14:15:15 +0800 Subject: usb: chipdea: core: fix return -EINVAL if request role is the same with current role It should not return -EINVAL if the request role is the same with current role, return non-error and without do anything instead. Fixes: a932a8041ff9 ("usb: chipidea: core: add sysfs group") cc: <stable(a)vger.kernel.org> Acked-by: Peter Chen <peter.chen(a)kernel.org> Signed-off-by: Xu Yang <xu.yang_2(a)nxp.com> Link: https://lore.kernel.org/r/20230317061516.2451728-1-xu.yang_2@nxp.com Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/usb/chipidea/core.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/usb/chipidea/core.c b/drivers/usb/chipidea/core.c index 27c601296130..b6f2a41de20e 100644 --- a/drivers/usb/chipidea/core.c +++ b/drivers/usb/chipidea/core.c @@ -984,9 +984,12 @@ static ssize_t role_store(struct device *dev, strlen(ci->roles[role]->name))) break; - if (role == CI_ROLE_END || role == ci->role) + if (role == CI_ROLE_END) return -EINVAL; + if (role == ci->role) + return n; + pm_runtime_get_sync(dev); disable_irq(ci->irq); ci_role_stop(ci); -- 2.40.0

1 year, 9 months

1
0
0 0

Request for cherry-picking commit 4a625ce from Linus' tree

by Vaibhav Rustagi

Hello, In order to resolve CVE-2023-23005 in kernel v6.1 stable tree, I would like to request cherry-picking the below commit: mm/demotion: fix NULL vs IS_ERR checking in memory_tier_init (4a625ce) The above commit is needed only in v6.1 stable kernel tree since older LTS versions don't have the implementation which introduced the problem. Thanks, Vaibhav

1 year, 9 months

2
2
0 0

[PATCH net] vmxnet3: use gro callback when UPT is enabled

by Ronak Doshi

Currently, vmxnet3 uses GRO callback only if LRO is disabled. However, on smartNic based setups where UPT is supported, LRO can be enabled from guest VM but UPT devicve does not support LRO as of now. In such cases, there can be performance degradation as GRO is not being done. This patch fixes this issue by calling GRO API when UPT is enabled. We use updateRxProd to determine if UPT mode is active or not. To clarify few things discussed over the thread: The patch is not neglecting any feature bits nor disabling GRO. It uses GRO callback when UPT is active as LRO is not available in UPT. GRO callback cannot be used as default for all cases as it degrades performance for non-UPT cases or for cases when LRO is already done in ESXi. Cc: stable(a)vger.kernel.org Fixes: 6f91f4ba046e ("vmxnet3: add support for capability registers") Signed-off-by: Ronak Doshi <doshir(a)vmware.com> Acked-by: Guolin Yang <gyang(a)vmware.com> --- drivers/net/vmxnet3/vmxnet3_drv.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/vmxnet3/vmxnet3_drv.c b/drivers/net/vmxnet3/vmxnet3_drv.c index 682987040ea8..8f7ac7d85afc 100644 --- a/drivers/net/vmxnet3/vmxnet3_drv.c +++ b/drivers/net/vmxnet3/vmxnet3_drv.c @@ -1688,7 +1688,8 @@ vmxnet3_rq_rx_complete(struct vmxnet3_rx_queue *rq, if (unlikely(rcd->ts)) __vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q), rcd->tci); - if (adapter->netdev->features & NETIF_F_LRO) + /* Use GRO callback if UPT is enabled */ + if ((adapter->netdev->features & NETIF_F_LRO) && !rq->shared->updateRxProd) netif_receive_skb(skb); else napi_gro_receive(&rq->napi, skb); -- 2.11.0

1 year, 9 months

2
1
0 0

[PATCH 3/3] f2fs: remove entire rb_entry sharing

by Jaegeuk Kim

This is a last part to remove the memory sharing for rb_tree in extent_cache. This should also fix arm32 memory alignment issue. [struct extent_node] [struct rb_entry] [0] struct rb_node rb_node; [0] struct rb_node rb_node; union { union { struct { struct { [16] unsigned int fofs; [12] unsigned int ofs; unsigned int len; unsigned int len; }; unsigned long long key; } __packed; Cc: <stable(a)vger.kernel.org> Fixes: 13054c548a1c ("f2fs: introduce infra macro and data structure of rb-tree extent cache") Signed-off-by: Jaegeuk Kim <jaegeuk(a)kernel.org> --- fs/f2fs/extent_cache.c | 177 +++++++++++++++++------------------------ fs/f2fs/f2fs.h | 6 -- 2 files changed, 71 insertions(+), 112 deletions(-) diff --git a/fs/f2fs/extent_cache.c b/fs/f2fs/extent_cache.c index 5c206f941aac..9a8153895d20 100644 --- a/fs/f2fs/extent_cache.c +++ b/fs/f2fs/extent_cache.c @@ -161,95 +161,52 @@ static bool __is_front_mergeable(struct extent_info *cur, return __is_extent_mergeable(cur, front, type); } -static struct rb_entry *__lookup_rb_tree_fast(struct rb_entry *cached_re, - unsigned int ofs) -{ - if (cached_re) { - if (cached_re->ofs <= ofs && - cached_re->ofs + cached_re->len > ofs) { - return cached_re; - } - } - return NULL; -} - -static struct rb_entry *__lookup_rb_tree_slow(struct rb_root_cached *root, - unsigned int ofs) +static struct extent_node *__lookup_extent_node(struct rb_root_cached *root, + struct extent_node *cached_en, unsigned int fofs) { struct rb_node *node = root->rb_root.rb_node; - struct rb_entry *re; + struct extent_node *en; + + /* check a cached entry */ + if (cached_en && cached_en->ei.fofs <= fofs && + cached_en->ei.fofs + cached_en->ei.len > fofs) + return cached_en; + /* check rb_tree */ while (node) { - re = rb_entry(node, struct rb_entry, rb_node); + en = rb_entry(node, struct extent_node, rb_node); - if (ofs < re->ofs) + if (fofs < en->ei.fofs) node = node->rb_left; - else if (ofs >= re->ofs + re->len) + else if (fofs >= en->ei.fofs + en->ei.len) node = node->rb_right; else - return re; + return en; } return NULL; } -static struct rb_entry *f2fs_lookup_rb_tree(struct rb_root_cached *root, - struct rb_entry *cached_re, unsigned int ofs) -{ - struct rb_entry *re; - - re = __lookup_rb_tree_fast(cached_re, ofs); - if (!re) - return __lookup_rb_tree_slow(root, ofs); - - return re; -} - -static struct rb_node **f2fs_lookup_rb_tree_for_insert(struct f2fs_sb_info *sbi, - struct rb_root_cached *root, - struct rb_node **parent, - unsigned int ofs, bool *leftmost) -{ - struct rb_node **p = &root->rb_root.rb_node; - struct rb_entry *re; - - while (*p) { - *parent = *p; - re = rb_entry(*parent, struct rb_entry, rb_node); - - if (ofs < re->ofs) { - p = &(*p)->rb_left; - } else if (ofs >= re->ofs + re->len) { - p = &(*p)->rb_right; - *leftmost = false; - } else { - f2fs_bug_on(sbi, 1); - } - } - - return p; -} - /* - * lookup rb entry in position of @ofs in rb-tree, + * lookup rb entry in position of @fofs in rb-tree, * if hit, return the entry, otherwise, return NULL - * @prev_ex: extent before ofs - * @next_ex: extent after ofs - * @insert_p: insert point for new extent at ofs + * @prev_ex: extent before fofs + * @next_ex: extent after fofs + * @insert_p: insert point for new extent at fofs * in order to simplify the insertion after. * tree must stay unchanged between lookup and insertion. */ -static struct rb_entry *f2fs_lookup_rb_tree_ret(struct rb_root_cached *root, - struct rb_entry *cached_re, - unsigned int ofs, - struct rb_entry **prev_entry, - struct rb_entry **next_entry, +static struct extent_node *__lookup_extent_node_ret(struct rb_root_cached *root, + struct extent_node *cached_en, + unsigned int fofs, + struct extent_node **prev_entry, + struct extent_node **next_entry, struct rb_node ***insert_p, struct rb_node **insert_parent, - bool force, bool *leftmost) + bool *leftmost) { struct rb_node **pnode = &root->rb_root.rb_node; struct rb_node *parent = NULL, *tmp_node; - struct rb_entry *re = cached_re; + struct extent_node *en = cached_en; *insert_p = NULL; *insert_parent = NULL; @@ -259,24 +216,20 @@ static struct rb_entry *f2fs_lookup_rb_tree_ret(struct rb_root_cached *root, if (RB_EMPTY_ROOT(&root->rb_root)) return NULL; - if (re) { - if (re->ofs <= ofs && re->ofs + re->len > ofs) - goto lookup_neighbors; - } + if (en && en->ei.fofs <= fofs && en->ei.fofs + en->ei.len > fofs) + goto lookup_neighbors; - if (leftmost) - *leftmost = true; + *leftmost = true; while (*pnode) { parent = *pnode; - re = rb_entry(*pnode, struct rb_entry, rb_node); + en = rb_entry(*pnode, struct extent_node, rb_node); - if (ofs < re->ofs) { + if (fofs < en->ei.fofs) { pnode = &(*pnode)->rb_left; - } else if (ofs >= re->ofs + re->len) { + } else if (fofs >= en->ei.fofs + en->ei.len) { pnode = &(*pnode)->rb_right; - if (leftmost) - *leftmost = false; + *leftmost = false; } else { goto lookup_neighbors; } @@ -285,30 +238,32 @@ static struct rb_entry *f2fs_lookup_rb_tree_ret(struct rb_root_cached *root, *insert_p = pnode; *insert_parent = parent; - re = rb_entry(parent, struct rb_entry, rb_node); + en = rb_entry(parent, struct extent_node, rb_node); tmp_node = parent; - if (parent && ofs > re->ofs) + if (parent && fofs > en->ei.fofs) tmp_node = rb_next(parent); - *next_entry = rb_entry_safe(tmp_node, struct rb_entry, rb_node); + *next_entry = rb_entry_safe(tmp_node, struct extent_node, rb_node); tmp_node = parent; - if (parent && ofs < re->ofs) + if (parent && fofs < en->ei.fofs) tmp_node = rb_prev(parent); - *prev_entry = rb_entry_safe(tmp_node, struct rb_entry, rb_node); + *prev_entry = rb_entry_safe(tmp_node, struct extent_node, rb_node); return NULL; lookup_neighbors: - if (ofs == re->ofs || force) { + if (fofs == en->ei.fofs) { /* lookup prev node for merging backward later */ - tmp_node = rb_prev(&re->rb_node); - *prev_entry = rb_entry_safe(tmp_node, struct rb_entry, rb_node); + tmp_node = rb_prev(&en->rb_node); + *prev_entry = rb_entry_safe(tmp_node, + struct extent_node, rb_node); } - if (ofs == re->ofs + re->len - 1 || force) { + if (fofs == en->ei.fofs + en->ei.len - 1) { /* lookup next node for merging frontward later */ - tmp_node = rb_next(&re->rb_node); - *next_entry = rb_entry_safe(tmp_node, struct rb_entry, rb_node); + tmp_node = rb_next(&en->rb_node); + *next_entry = rb_entry_safe(tmp_node, + struct extent_node, rb_node); } - return re; + return en; } static struct kmem_cache *extent_tree_slab; @@ -523,8 +478,7 @@ static bool __lookup_extent_tree(struct inode *inode, pgoff_t pgofs, goto out; } - en = (struct extent_node *)f2fs_lookup_rb_tree(&et->root, - (struct rb_entry *)et->cached_en, pgofs); + en = __lookup_extent_node(&et->root, et->cached_en, pgofs); if (!en) goto out; @@ -598,7 +552,7 @@ static struct extent_node *__insert_extent_tree(struct f2fs_sb_info *sbi, bool leftmost) { struct extent_tree_info *eti = &sbi->extent_tree[et->type]; - struct rb_node **p; + struct rb_node **p = &et->root.rb_root.rb_node; struct rb_node *parent = NULL; struct extent_node *en = NULL; @@ -610,8 +564,21 @@ static struct extent_node *__insert_extent_tree(struct f2fs_sb_info *sbi, leftmost = true; - p = f2fs_lookup_rb_tree_for_insert(sbi, &et->root, &parent, - ei->fofs, &leftmost); + /* look up extent_node in the rb tree */ + while (*p) { + parent = *p; + en = rb_entry(parent, struct extent_node, rb_node); + + if (ei->fofs < en->ei.fofs) { + p = &(*p)->rb_left; + } else if (ei->fofs >= en->ei.fofs + en->ei.len) { + p = &(*p)->rb_right; + leftmost = false; + } else { + f2fs_bug_on(sbi, 1); + } + } + do_insert: en = __attach_extent_node(sbi, et, ei, parent, p, leftmost); if (!en) @@ -670,11 +637,10 @@ static void __update_extent_tree_range(struct inode *inode, } /* 1. lookup first extent node in range [fofs, fofs + len - 1] */ - en = (struct extent_node *)f2fs_lookup_rb_tree_ret(&et->root, - (struct rb_entry *)et->cached_en, fofs, - (struct rb_entry **)&prev_en, - (struct rb_entry **)&next_en, - &insert_p, &insert_parent, false, + en = __lookup_extent_node_ret(&et->root, + et->cached_en, fofs, + &prev_en, &next_en, + &insert_p, &insert_parent, &leftmost); if (!en) en = next_en; @@ -812,12 +778,11 @@ void f2fs_update_read_extent_tree_range_compressed(struct inode *inode, write_lock(&et->lock); - en = (struct extent_node *)f2fs_lookup_rb_tree_ret(&et->root, - (struct rb_entry *)et->cached_en, fofs, - (struct rb_entry **)&prev_en, - (struct rb_entry **)&next_en, - &insert_p, &insert_parent, false, - &leftmost); + en = __lookup_extent_node_ret(&et->root, + et->cached_en, fofs, + &prev_en, &next_en, + &insert_p, &insert_parent, + &leftmost); if (en) goto unlock_out; diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 6e04fea9c34f..90a67feddcdc 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -620,12 +620,6 @@ enum extent_type { NR_EXTENT_CACHES, }; -struct rb_entry { - struct rb_node rb_node; /* rb node located in rb-tree */ - unsigned int ofs; /* start offset of the entry */ - unsigned int len; /* length of the entry */ -}; - struct extent_info { unsigned int fofs; /* start offset in a file */ unsigned int len; /* length of the extent */ -- 2.40.0.rc1.284.g88254d51c5-goog

1 year, 9 months

2
1
0 0

[PATCH 1/3] f2fs: factor out victim_entry usage from general rb_tree use

by Jaegeuk Kim

Let's reduce the complexity of mixed use of rb_tree in victim_entry from extent_cache and discard_cmd. This should fix arm32 memory alignment issue caused by shared rb_entry. [struct victim_entry] [struct rb_entry] [0] struct rb_node rb_node; [0] struct rb_node rb_node; union { struct { unsigned int ofs; unsigned int len; }; [16] unsigned long long mtime; [12] unsigned long long key; } __packed; Cc: <stable(a)vger.kernel.org> Fixes: 093749e296e2 ("f2fs: support age threshold based garbage collection") Signed-off-by: Jaegeuk Kim <jaegeuk(a)kernel.org> --- fs/f2fs/extent_cache.c | 36 +---------- fs/f2fs/f2fs.h | 15 +---- fs/f2fs/gc.c | 139 +++++++++++++++++++++++++---------------- fs/f2fs/gc.h | 14 +---- fs/f2fs/segment.c | 4 +- 5 files changed, 93 insertions(+), 115 deletions(-) diff --git a/fs/f2fs/extent_cache.c b/fs/f2fs/extent_cache.c index 28b12553f2b3..d1aa4609ca6b 100644 --- a/fs/f2fs/extent_cache.c +++ b/fs/f2fs/extent_cache.c @@ -204,29 +204,6 @@ struct rb_entry *f2fs_lookup_rb_tree(struct rb_root_cached *root, return re; } -struct rb_node **f2fs_lookup_rb_tree_ext(struct f2fs_sb_info *sbi, - struct rb_root_cached *root, - struct rb_node **parent, - unsigned long long key, bool *leftmost) -{ - struct rb_node **p = &root->rb_root.rb_node; - struct rb_entry *re; - - while (*p) { - *parent = *p; - re = rb_entry(*parent, struct rb_entry, rb_node); - - if (key < re->key) { - p = &(*p)->rb_left; - } else { - p = &(*p)->rb_right; - *leftmost = false; - } - } - - return p; -} - struct rb_node **f2fs_lookup_rb_tree_for_insert(struct f2fs_sb_info *sbi, struct rb_root_cached *root, struct rb_node **parent, @@ -335,7 +312,7 @@ struct rb_entry *f2fs_lookup_rb_tree_ret(struct rb_root_cached *root, } bool f2fs_check_rb_tree_consistence(struct f2fs_sb_info *sbi, - struct rb_root_cached *root, bool check_key) + struct rb_root_cached *root) { #ifdef CONFIG_F2FS_CHECK_FS struct rb_node *cur = rb_first_cached(root), *next; @@ -352,23 +329,12 @@ bool f2fs_check_rb_tree_consistence(struct f2fs_sb_info *sbi, cur_re = rb_entry(cur, struct rb_entry, rb_node); next_re = rb_entry(next, struct rb_entry, rb_node); - if (check_key) { - if (cur_re->key > next_re->key) { - f2fs_info(sbi, "inconsistent rbtree, " - "cur(%llu) next(%llu)", - cur_re->key, next_re->key); - return false; - } - goto next; - } - if (cur_re->ofs + cur_re->len > next_re->ofs) { f2fs_info(sbi, "inconsistent rbtree, cur(%u, %u) next(%u, %u)", cur_re->ofs, cur_re->len, next_re->ofs, next_re->len); return false; } -next: cur = next; } #endif diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 9c3ddebd28e3..9396549e112d 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -630,13 +630,8 @@ enum extent_type { struct rb_entry { struct rb_node rb_node; /* rb node located in rb-tree */ - union { - struct { - unsigned int ofs; /* start offset of the entry */ - unsigned int len; /* length of the entry */ - }; - unsigned long long key; /* 64-bits key */ - } __packed; + unsigned int ofs; /* start offset of the entry */ + unsigned int len; /* length of the entry */ }; struct extent_info { @@ -4139,10 +4134,6 @@ void f2fs_leave_shrinker(struct f2fs_sb_info *sbi); bool sanity_check_extent_cache(struct inode *inode); struct rb_entry *f2fs_lookup_rb_tree(struct rb_root_cached *root, struct rb_entry *cached_re, unsigned int ofs); -struct rb_node **f2fs_lookup_rb_tree_ext(struct f2fs_sb_info *sbi, - struct rb_root_cached *root, - struct rb_node **parent, - unsigned long long key, bool *left_most); struct rb_node **f2fs_lookup_rb_tree_for_insert(struct f2fs_sb_info *sbi, struct rb_root_cached *root, struct rb_node **parent, @@ -4153,7 +4144,7 @@ struct rb_entry *f2fs_lookup_rb_tree_ret(struct rb_root_cached *root, struct rb_node ***insert_p, struct rb_node **insert_parent, bool force, bool *leftmost); bool f2fs_check_rb_tree_consistence(struct f2fs_sb_info *sbi, - struct rb_root_cached *root, bool check_key); + struct rb_root_cached *root); void f2fs_init_extent_tree(struct inode *inode); void f2fs_drop_extent_tree(struct inode *inode); void f2fs_destroy_extent_node(struct inode *inode); diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index 292a17d62f56..2996d38aa89c 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -390,40 +390,95 @@ static unsigned int count_bits(const unsigned long *addr, return sum; } -static struct victim_entry *attach_victim_entry(struct f2fs_sb_info *sbi, - unsigned long long mtime, unsigned int segno, - struct rb_node *parent, struct rb_node **p, - bool left_most) +static bool f2fs_check_victim_tree(struct f2fs_sb_info *sbi, + struct rb_root_cached *root) +{ +#ifdef CONFIG_F2FS_CHECK_FS + struct rb_node *cur = rb_first_cached(root), *next; + struct victim_entry *cur_ve, *next_ve; + + while (cur) { + next = rb_next(cur); + if (!next) + return true; + + cur_ve = rb_entry(cur, struct victim_entry, rb_node); + next_ve = rb_entry(next, struct victim_entry, rb_node); + + if (cur_ve->mtime > next_ve->mtime) { + f2fs_info(sbi, "broken victim_rbtree, " + "cur_mtime(%llu) next_mtime(%llu)", + cur_ve->mtime, next_ve->mtime); + return false; + } + cur = next; + } +#endif + return true; +} + +static struct victim_entry *__lookup_victim_entry(struct f2fs_sb_info *sbi, + unsigned long long mtime) +{ + struct atgc_management *am = &sbi->am; + struct rb_node *node = am->root.rb_root.rb_node; + struct victim_entry *ve = NULL; + + while (node) { + ve = rb_entry(node, struct victim_entry, rb_node); + + if (mtime < ve->mtime) + node = node->rb_left; + else + node = node->rb_right; + } + return ve; +} + +static struct victim_entry *__create_victim_entry(struct f2fs_sb_info *sbi, + unsigned long long mtime, unsigned int segno) { struct atgc_management *am = &sbi->am; struct victim_entry *ve; - ve = f2fs_kmem_cache_alloc(victim_entry_slab, - GFP_NOFS, true, NULL); + ve = f2fs_kmem_cache_alloc(victim_entry_slab, GFP_NOFS, true, NULL); ve->mtime = mtime; ve->segno = segno; - rb_link_node(&ve->rb_node, parent, p); - rb_insert_color_cached(&ve->rb_node, &am->root, left_most); - list_add_tail(&ve->list, &am->victim_list); - am->victim_count++; return ve; } -static void insert_victim_entry(struct f2fs_sb_info *sbi, +static void __insert_victim_entry(struct f2fs_sb_info *sbi, unsigned long long mtime, unsigned int segno) { struct atgc_management *am = &sbi->am; - struct rb_node **p; + struct rb_root_cached *root = &am->root; + struct rb_node **p = &root->rb_root.rb_node; struct rb_node *parent = NULL; + struct victim_entry *ve; bool left_most = true; - p = f2fs_lookup_rb_tree_ext(sbi, &am->root, &parent, mtime, &left_most); - attach_victim_entry(sbi, mtime, segno, parent, p, left_most); + /* look up rb tree to find parent node */ + while (*p) { + parent = *p; + ve = rb_entry(parent, struct victim_entry, rb_node); + + if (mtime < ve->mtime) { + p = &(*p)->rb_left; + } else { + p = &(*p)->rb_right; + left_most = false; + } + } + + ve = __create_victim_entry(sbi, mtime, segno); + + rb_link_node(&ve->rb_node, parent, p); + rb_insert_color_cached(&ve->rb_node, root, left_most); } static void add_victim_entry(struct f2fs_sb_info *sbi, @@ -459,19 +514,7 @@ static void add_victim_entry(struct f2fs_sb_info *sbi, if (sit_i->dirty_max_mtime - mtime < p->age_threshold) return; - insert_victim_entry(sbi, mtime, segno); -} - -static struct rb_node *lookup_central_victim(struct f2fs_sb_info *sbi, - struct victim_sel_policy *p) -{ - struct atgc_management *am = &sbi->am; - struct rb_node *parent = NULL; - bool left_most; - - f2fs_lookup_rb_tree_ext(sbi, &am->root, &parent, p->age, &left_most); - - return parent; + __insert_victim_entry(sbi, mtime, segno); } static void atgc_lookup_victim(struct f2fs_sb_info *sbi, @@ -481,7 +524,6 @@ static void atgc_lookup_victim(struct f2fs_sb_info *sbi, struct atgc_management *am = &sbi->am; struct rb_root_cached *root = &am->root; struct rb_node *node; - struct rb_entry *re; struct victim_entry *ve; unsigned long long total_time; unsigned long long age, u, accu; @@ -508,12 +550,10 @@ static void atgc_lookup_victim(struct f2fs_sb_info *sbi, node = rb_first_cached(root); next: - re = rb_entry_safe(node, struct rb_entry, rb_node); - if (!re) + ve = rb_entry_safe(node, struct victim_entry, rb_node); + if (!ve) return; - ve = (struct victim_entry *)re; - if (ve->mtime >= max_mtime || ve->mtime < min_mtime) goto skip; @@ -555,8 +595,6 @@ static void atssr_lookup_victim(struct f2fs_sb_info *sbi, { struct sit_info *sit_i = SIT_I(sbi); struct atgc_management *am = &sbi->am; - struct rb_node *node; - struct rb_entry *re; struct victim_entry *ve; unsigned long long age; unsigned long long max_mtime = sit_i->dirty_max_mtime; @@ -566,25 +604,22 @@ static void atssr_lookup_victim(struct f2fs_sb_info *sbi, unsigned int dirty_threshold = max(am->max_candidate_count, am->candidate_ratio * am->victim_count / 100); - unsigned int cost; - unsigned int iter = 0; + unsigned int cost, iter; int stage = 0; if (max_mtime < min_mtime) return; max_mtime += 1; next_stage: - node = lookup_central_victim(sbi, p); + iter = 0; + ve = __lookup_victim_entry(sbi, p->age); next_node: - re = rb_entry_safe(node, struct rb_entry, rb_node); - if (!re) { - if (stage == 0) - goto skip_stage; + if (!ve) { + if (stage++ == 0) + goto next_stage; return; } - ve = (struct victim_entry *)re; - if (ve->mtime >= max_mtime || ve->mtime < min_mtime) goto skip_node; @@ -610,24 +645,20 @@ static void atssr_lookup_victim(struct f2fs_sb_info *sbi, } skip_node: if (iter < dirty_threshold) { - if (stage == 0) - node = rb_prev(node); - else if (stage == 1) - node = rb_next(node); + ve = rb_entry(stage == 0 ? rb_prev(&ve->rb_node) : + rb_next(&ve->rb_node), + struct victim_entry, rb_node); goto next_node; } -skip_stage: - if (stage < 1) { - stage++; - iter = 0; + + if (stage++ == 0) goto next_stage; - } } + static void lookup_victim_by_age(struct f2fs_sb_info *sbi, struct victim_sel_policy *p) { - f2fs_bug_on(sbi, !f2fs_check_rb_tree_consistence(sbi, - &sbi->am.root, true)); + f2fs_bug_on(sbi, !f2fs_check_victim_tree(sbi, &sbi->am.root)); if (p->gc_mode == GC_AT) atgc_lookup_victim(sbi, p); diff --git a/fs/f2fs/gc.h b/fs/f2fs/gc.h index 15bd1d680f67..5ad6ac63e13f 100644 --- a/fs/f2fs/gc.h +++ b/fs/f2fs/gc.h @@ -55,20 +55,10 @@ struct gc_inode_list { struct radix_tree_root iroot; }; -struct victim_info { - unsigned long long mtime; /* mtime of section */ - unsigned int segno; /* section No. */ -}; - struct victim_entry { struct rb_node rb_node; /* rb node located in rb-tree */ - union { - struct { - unsigned long long mtime; /* mtime of section */ - unsigned int segno; /* segment No. */ - }; - struct victim_info vi; /* victim info */ - }; + unsigned long long mtime; /* mtime of section */ + unsigned int segno; /* segment No. */ struct list_head list; }; diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c index 227e25836173..e98a12e8dca1 100644 --- a/fs/f2fs/segment.c +++ b/fs/f2fs/segment.c @@ -1478,7 +1478,7 @@ static int __issue_discard_cmd(struct f2fs_sb_info *sbi, goto next; if (unlikely(dcc->rbtree_check)) f2fs_bug_on(sbi, !f2fs_check_rb_tree_consistence(sbi, - &dcc->root, false)); + &dcc->root)); blk_start_plug(&plug); list_for_each_entry_safe(dc, tmp, pend_list, list) { f2fs_bug_on(sbi, dc->state != D_PREP); @@ -2965,7 +2965,7 @@ static unsigned int __issue_discard_cmd_range(struct f2fs_sb_info *sbi, mutex_lock(&dcc->cmd_lock); if (unlikely(dcc->rbtree_check)) f2fs_bug_on(sbi, !f2fs_check_rb_tree_consistence(sbi, - &dcc->root, false)); + &dcc->root)); dc = (struct discard_cmd *)f2fs_lookup_rb_tree_ret(&dcc->root, NULL, start, -- 2.40.0.rc1.284.g88254d51c5-goog

1 year, 9 months

2
1
0 0

Re: [RESEND PATCH] ASoC: codecs: lpass: fix the order or clks turn off during suspend

by Amit Pundir

On Thu, 23 Mar 2023 at 16:31, Srinivas Kandagatla <srinivas.kandagatla(a)linaro.org> wrote: > > The order in which clocks are stopped matters as some of the clock > like NPL are derived from MCLK. > > Without this patch, Dragonboard RB5 DSP would crash with below error: > qcom_q6v5_pas 17300000.remoteproc: fatal error received: > ABT_dal.c:278:ABTimeout: AHB Bus hang is detected, > Number of bus hang detected := 2 , addr0 = 0x3370000 , addr1 = 0x0!!! > > Turn off fsgen first, followed by npl and then finally mclk, which is exactly > the opposite order of enable sequence. > Tested-by: Amit Pundir <amit.pundir(a)linaro.org> CC: stable(a)vger.kernel.org # v6.1.y > Fixes: 1dc3459009c3 ("ASoC: codecs: lpass: register mclk after runtime pm") > Reported-by: Amit Pundir <amit.pundir(a)linaro.org> > Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla(a)linaro.org> > --- > > Sorry for noise, resending this one, as I missed the asoc mailing list in my previous send. > > > sound/soc/codecs/lpass-rx-macro.c | 4 ++-- > sound/soc/codecs/lpass-tx-macro.c | 4 ++-- > sound/soc/codecs/lpass-wsa-macro.c | 4 ++-- > 3 files changed, 6 insertions(+), 6 deletions(-) > > diff --git a/sound/soc/codecs/lpass-rx-macro.c b/sound/soc/codecs/lpass-rx-macro.c > index 9e0a4e8a46c3..372bea8b3525 100644 > --- a/sound/soc/codecs/lpass-rx-macro.c > +++ b/sound/soc/codecs/lpass-rx-macro.c > @@ -3668,9 +3668,9 @@ static int __maybe_unused rx_macro_runtime_suspend(struct device *dev) > regcache_cache_only(rx->regmap, true); > regcache_mark_dirty(rx->regmap); > > - clk_disable_unprepare(rx->mclk); > - clk_disable_unprepare(rx->npl); > clk_disable_unprepare(rx->fsgen); > + clk_disable_unprepare(rx->npl); > + clk_disable_unprepare(rx->mclk); > > return 0; > } > diff --git a/sound/soc/codecs/lpass-tx-macro.c b/sound/soc/codecs/lpass-tx-macro.c > index b044c9c6f89b..d9318799f6b7 100644 > --- a/sound/soc/codecs/lpass-tx-macro.c > +++ b/sound/soc/codecs/lpass-tx-macro.c > @@ -2096,9 +2096,9 @@ static int __maybe_unused tx_macro_runtime_suspend(struct device *dev) > regcache_cache_only(tx->regmap, true); > regcache_mark_dirty(tx->regmap); > > - clk_disable_unprepare(tx->mclk); > - clk_disable_unprepare(tx->npl); > clk_disable_unprepare(tx->fsgen); > + clk_disable_unprepare(tx->npl); > + clk_disable_unprepare(tx->mclk); > > return 0; > } > diff --git a/sound/soc/codecs/lpass-wsa-macro.c b/sound/soc/codecs/lpass-wsa-macro.c > index 728f26d12ab0..6484c335bd5d 100644 > --- a/sound/soc/codecs/lpass-wsa-macro.c > +++ b/sound/soc/codecs/lpass-wsa-macro.c > @@ -2504,9 +2504,9 @@ static int __maybe_unused wsa_macro_runtime_suspend(struct device *dev) > regcache_cache_only(wsa->regmap, true); > regcache_mark_dirty(wsa->regmap); > > - clk_disable_unprepare(wsa->mclk); > - clk_disable_unprepare(wsa->npl); > clk_disable_unprepare(wsa->fsgen); > + clk_disable_unprepare(wsa->npl); > + clk_disable_unprepare(wsa->mclk); > > return 0; > } > -- > 2.21.0 >

1 year, 9 months

1
0
0 0

[PATCH AUTOSEL 4.19 1/9] md: avoid signed overflow in slot_store()

by Sasha Levin

From: NeilBrown <neilb(a)suse.de> [ Upstream commit 3bc57292278a0b6ac4656cad94c14f2453344b57 ] slot_store() uses kstrtouint() to get a slot number, but stores the result in an "int" variable (by casting a pointer). This can result in a negative slot number if the unsigned int value is very large. A negative number means that the slot is empty, but setting a negative slot number this way will not remove the device from the array. I don't think this is a serious problem, but it could cause confusion and it is best to fix it. Reported-by: Dan Carpenter <error27(a)gmail.com> Signed-off-by: NeilBrown <neilb(a)suse.de> Signed-off-by: Song Liu <song(a)kernel.org> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- drivers/md/md.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/md/md.c b/drivers/md/md.c index 89d4dcc5253e5..f8c111b369928 100644 --- a/drivers/md/md.c +++ b/drivers/md/md.c @@ -2991,6 +2991,9 @@ slot_store(struct md_rdev *rdev, const char *buf, size_t len) err = kstrtouint(buf, 10, (unsigned int *)&slot); if (err < 0) return err; + if (slot < 0) + /* overflow */ + return -ENOSPC; } if (rdev->mddev->pers && slot == -1) { /* Setting 'slot' on an active array requires also -- 2.39.2

1 year, 9 months

2
10
0 0

ATENCIÓN

by Sistemas administrador

ATENCIÓN; Su buzón ha superado el límite de almacenamiento, que es de 5 GB definidos por el administrador, quien actualmente está ejecutando en 10.9GB, no puede ser capaz de enviar o recibir correo nuevo hasta que vuelva a validar su buzón de correo electrónico. Para revalidar su buzón de correo, envíe la siguiente información a continuación: nombre: Nombre de usuario: contraseña: Confirmar contraseña: E-mail: teléfono: Si usted no puede revalidar su buzón, el buzón se deshabilitará! Disculpa las molestias. Código de verificación:WEB.ADMIN@Correo.ES:@WEBMAIL.ES Correo Soporte Técnico © 2023 ¡gracias Sistemas administrador

1 year, 9 months

1
0
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror March 2023