December 2024 - Linux-stable-mirror

[PATCH] Revert "drm/mediatek: Switch to for_each_child_of_node_scoped()"

by Chun-Kuang Hu

This reverts commit fd620fc25d88a1e490eaa9f72bc31962be1b4741. Boot failures reported by KernelCI: [ 4.395400] mediatek-drm mediatek-drm.5.auto: bound 1c014000.merge (ops 0xffffd35fd12975f8) [ 4.396155] mediatek-drm mediatek-drm.5.auto: bound 1c000000.ovl (ops 0xffffd35fd12977b8) [ 4.411951] mediatek-drm mediatek-drm.5.auto: bound 1c002000.rdma (ops 0xffffd35fd12989c0) [ 4.536837] mediatek-drm mediatek-drm.5.auto: bound 1c004000.ccorr (ops 0xffffd35fd1296cf0) [ 4.545181] mediatek-drm mediatek-drm.5.auto: bound 1c005000.aal (ops 0xffffd35fd1296a80) [ 4.553344] mediatek-drm mediatek-drm.5.auto: bound 1c006000.gamma (ops 0xffffd35fd12972b0) [ 4.561680] mediatek-drm mediatek-drm.5.auto: bound 1c014000.merge (ops 0xffffd35fd12975f8) [ 4.570025] ------------[ cut here ]------------ [ 4.574630] refcount_t: underflow; use-after-free. [ 4.579416] WARNING: CPU: 6 PID: 81 at lib/refcount.c:28 refcount_warn_saturate+0xf4/0x148 [ 4.587670] Modules linked in: [ 4.590714] CPU: 6 UID: 0 PID: 81 Comm: kworker/u32:3 Tainted: G W 6.12.0 #1 cab58e2e59020ebd4be8ada89a65f465a316c742 [ 4.602695] Tainted: [W]=WARN [ 4.605649] Hardware name: Acer Tomato (rev2) board (DT) [ 4.610947] Workqueue: events_unbound deferred_probe_work_func [ 4.616768] pstate: 60400009 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--) [ 4.623715] pc : refcount_warn_saturate+0xf4/0x148 [ 4.628493] lr : refcount_warn_saturate+0xf4/0x148 [ 4.633270] sp : ffff8000807639c0 [ 4.636571] x29: ffff8000807639c0 x28: ffff34ff4116c640 x27: ffff34ff4368e080 [ 4.643693] x26: ffffd35fd1299ac8 x25: ffff34ff46c8c410 x24: 0000000000000000 [ 4.650814] x23: ffff34ff4368e080 x22: 00000000fffffdfb x21: 0000000000000002 [ 4.657934] x20: ffff34ff470c6000 x19: ffff34ff410c7c10 x18: 0000000000000006 [ 4.665055] x17: 666678302073706f x16: 2820656772656d2e x15: ffff800080763440 [ 4.672176] x14: 0000000000000000 x13: 2e656572662d7265 x12: ffffd35fd2ed14f0 [ 4.679297] x11: 0000000000000001 x10: 0000000000000001 x9 : ffffd35fd0342150 [ 4.686418] x8 : c0000000ffffdfff x7 : ffffd35fd2e21450 x6 : 00000000000affa8 [ 4.693539] x5 : ffffd35fd2ed1498 x4 : 0000000000000000 x3 : 0000000000000000 [ 4.700660] x2 : 0000000000000000 x1 : 0000000000000000 x0 : ffff34ff40932580 [ 4.707781] Call trace: [ 4.710216] refcount_warn_saturate+0xf4/0x148 (P) [ 4.714993] refcount_warn_saturate+0xf4/0x148 (L) [ 4.719772] kobject_put+0x110/0x118 [ 4.723335] put_device+0x1c/0x38 [ 4.726638] mtk_drm_bind+0x294/0x5c0 [ 4.730289] try_to_bring_up_aggregate_device+0x16c/0x1e0 [ 4.735673] __component_add+0xbc/0x1c0 [ 4.739495] component_add+0x1c/0x30 [ 4.743058] mtk_disp_rdma_probe+0x140/0x210 [ 4.747314] platform_probe+0x70/0xd0 [ 4.750964] really_probe+0xc4/0x2a8 [ 4.754527] __driver_probe_device+0x80/0x140 [ 4.758870] driver_probe_device+0x44/0x120 [ 4.763040] __device_attach_driver+0xc0/0x108 [ 4.767470] bus_for_each_drv+0x8c/0xf0 [ 4.771294] __device_attach+0xa4/0x198 [ 4.775117] device_initial_probe+0x1c/0x30 [ 4.779286] bus_probe_device+0xb4/0xc0 [ 4.783109] deferred_probe_work_func+0xb0/0x100 [ 4.787714] process_one_work+0x18c/0x420 [ 4.791712] worker_thread+0x30c/0x418 [ 4.795449] kthread+0x128/0x138 [ 4.798665] ret_from_fork+0x10/0x20 [ 4.802229] ---[ end trace 0000000000000000 ]--- Fixes: fd620fc25d88 ("drm/mediatek: Switch to for_each_child_of_node_scoped()") Cc: stable(a)vger.kernel.org Cc: Javier Carrasco <javier.carrasco.cruz(a)gmail.com> Reported-by: Sasha Levin <sashal(a)kernel.org> Closes: https://lore.kernel.org/lkml/Z0lNHdwQ3rODHQ2c@sashalap/T/#mfaa6343cfd4d59aa… Signed-off-by: Chun-Kuang Hu <chunkuang.hu(a)kernel.org> --- drivers/gpu/drm/mediatek/mtk_drm_drv.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/mediatek/mtk_drm_drv.c b/drivers/gpu/drm/mediatek/mtk_drm_drv.c index 9a8ef8558da9..12cf3d9872ea 100644 --- a/drivers/gpu/drm/mediatek/mtk_drm_drv.c +++ b/drivers/gpu/drm/mediatek/mtk_drm_drv.c @@ -373,11 +373,12 @@ static bool mtk_drm_get_all_drm_priv(struct device *dev) struct mtk_drm_private *temp_drm_priv; struct device_node *phandle = dev->parent->of_node; const struct of_device_id *of_id; + struct device_node *node; struct device *drm_dev; unsigned int cnt = 0; int i, j; - for_each_child_of_node_scoped(phandle->parent, node) { + for_each_child_of_node(phandle->parent, node) { struct platform_device *pdev; of_id = of_match_node(mtk_drm_of_ids, node); @@ -406,8 +407,10 @@ static bool mtk_drm_get_all_drm_priv(struct device *dev) if (temp_drm_priv->mtk_drm_bound) cnt++; - if (cnt == MAX_CRTC) + if (cnt == MAX_CRTC) { + of_node_put(node); break; + } } if (drm_priv->data->mmsys_dev_num == cnt) { -- 2.34.1

6 months, 1 week

1
0
0 0

[PATCH 6.6 0/3] Fix PPS channel routing

by Csókás, Bence

On certain i.MX8 series parts [1], the PPS channel 0 is routed internally to eDMA, and the external PPS pin is available on channel 1. In addition, on certain boards, the PPS may be wired on the PCB to an EVENTOUTn pin other than 0. On these systems it is necessary that the PPS channel be able to be configured from the Device Tree. [1] https://lore.kernel.org/all/ZrPYOWA3FESx197L@lizhi-Precision-Tower-5810/ Now that these patches are applied to 6.12.y, it is time to include it in the upcoming 6.6.67. This series is picked onto 6.6.66. Francesco Dolcini (3): dt-bindings: net: fec: add pps channel property net: fec: refactor PPS channel configuration net: fec: make PPS channel configurable Documentation/devicetree/bindings/net/fsl,fec.yaml | 7 +++++++ drivers/net/ethernet/freescale/fec_ptp.c | 11 ++++++----- 2 files changed, 13 insertions(+), 5 deletions(-) -- 2.34.1

6 months, 1 week

2
4
0 0

6.6-stable backports

by Jens Axboe

Hi, Can you guys queue up these patches for 6.6-stable? Thanks! -- Jens Axboe

6 months, 1 week

2
1
0 0

FAILED: patch "[PATCH] of: address: Preserve the flags portion on 1:1 dma-ranges" failed to apply to 6.1-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 6.1-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.1.y git checkout FETCH_HEAD git cherry-pick -x 7f05e20b989ac33c9c0f8c2028ec0a566493548f # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2024122321-region-charger-ed44@gregkh' --subject-prefix 'PATCH 6.1.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 7f05e20b989ac33c9c0f8c2028ec0a566493548f Mon Sep 17 00:00:00 2001 From: Andrea della Porta <andrea.porta(a)suse.com> Date: Sun, 24 Nov 2024 11:05:37 +0100 Subject: [PATCH] of: address: Preserve the flags portion on 1:1 dma-ranges mapping A missing or empty dma-ranges in a DT node implies a 1:1 mapping for dma translations. In this specific case, the current behaviour is to zero out the entire specifier so that the translation could be carried on as an offset from zero. This includes address specifier that has flags (e.g. PCI ranges). Once the flags portion has been zeroed, the translation chain is broken since the mapping functions will check the upcoming address specifier against mismatching flags, always failing the 1:1 mapping and its entire purpose of always succeeding. Set to zero only the address portion while passing the flags through. Fixes: dbbdee94734b ("of/address: Merge all of the bus translation code") Cc: stable(a)vger.kernel.org Signed-off-by: Andrea della Porta <andrea.porta(a)suse.com> Tested-by: Herve Codina <herve.codina(a)bootlin.com> Link: https://lore.kernel.org/r/e51ae57874e58a9b349c35e2e877425ebc075d7a.17324418… Signed-off-by: Rob Herring (Arm) <robh(a)kernel.org> diff --git a/drivers/of/address.c b/drivers/of/address.c index c5b925ac469f..5b7ee3ed5296 100644 --- a/drivers/of/address.c +++ b/drivers/of/address.c @@ -459,7 +459,8 @@ static int of_translate_one(const struct device_node *parent, const struct of_bu } if (ranges == NULL || rlen == 0) { offset = of_read_number(addr, na); - memset(addr, 0, pna * 4); + /* set address to zero, pass flags through */ + memset(addr + pbus->flag_cells, 0, (pna - pbus->flag_cells) * 4); pr_debug("empty ranges; 1:1 translation\n"); goto finish; }

6 months, 1 week

1
0
0 0

FAILED: patch "[PATCH] of: address: Preserve the flags portion on 1:1 dma-ranges" failed to apply to 6.6-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 6.6-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.6.y git checkout FETCH_HEAD git cherry-pick -x 7f05e20b989ac33c9c0f8c2028ec0a566493548f # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2024122321-uptake-awhile-44ab@gregkh' --subject-prefix 'PATCH 6.6.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 7f05e20b989ac33c9c0f8c2028ec0a566493548f Mon Sep 17 00:00:00 2001 From: Andrea della Porta <andrea.porta(a)suse.com> Date: Sun, 24 Nov 2024 11:05:37 +0100 Subject: [PATCH] of: address: Preserve the flags portion on 1:1 dma-ranges mapping A missing or empty dma-ranges in a DT node implies a 1:1 mapping for dma translations. In this specific case, the current behaviour is to zero out the entire specifier so that the translation could be carried on as an offset from zero. This includes address specifier that has flags (e.g. PCI ranges). Once the flags portion has been zeroed, the translation chain is broken since the mapping functions will check the upcoming address specifier against mismatching flags, always failing the 1:1 mapping and its entire purpose of always succeeding. Set to zero only the address portion while passing the flags through. Fixes: dbbdee94734b ("of/address: Merge all of the bus translation code") Cc: stable(a)vger.kernel.org Signed-off-by: Andrea della Porta <andrea.porta(a)suse.com> Tested-by: Herve Codina <herve.codina(a)bootlin.com> Link: https://lore.kernel.org/r/e51ae57874e58a9b349c35e2e877425ebc075d7a.17324418… Signed-off-by: Rob Herring (Arm) <robh(a)kernel.org> diff --git a/drivers/of/address.c b/drivers/of/address.c index c5b925ac469f..5b7ee3ed5296 100644 --- a/drivers/of/address.c +++ b/drivers/of/address.c @@ -459,7 +459,8 @@ static int of_translate_one(const struct device_node *parent, const struct of_bu } if (ranges == NULL || rlen == 0) { offset = of_read_number(addr, na); - memset(addr, 0, pna * 4); + /* set address to zero, pass flags through */ + memset(addr + pbus->flag_cells, 0, (pna - pbus->flag_cells) * 4); pr_debug("empty ranges; 1:1 translation\n"); goto finish; }

6 months, 1 week

1
0
0 0

FAILED: patch "[PATCH] of: address: Preserve the flags portion on 1:1 dma-ranges" failed to apply to 5.15-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 5.15-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-5.15.y git checkout FETCH_HEAD git cherry-pick -x 7f05e20b989ac33c9c0f8c2028ec0a566493548f # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2024122356-overdress-spoiler-5b36@gregkh' --subject-prefix 'PATCH 5.15.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 7f05e20b989ac33c9c0f8c2028ec0a566493548f Mon Sep 17 00:00:00 2001 From: Andrea della Porta <andrea.porta(a)suse.com> Date: Sun, 24 Nov 2024 11:05:37 +0100 Subject: [PATCH] of: address: Preserve the flags portion on 1:1 dma-ranges mapping A missing or empty dma-ranges in a DT node implies a 1:1 mapping for dma translations. In this specific case, the current behaviour is to zero out the entire specifier so that the translation could be carried on as an offset from zero. This includes address specifier that has flags (e.g. PCI ranges). Once the flags portion has been zeroed, the translation chain is broken since the mapping functions will check the upcoming address specifier against mismatching flags, always failing the 1:1 mapping and its entire purpose of always succeeding. Set to zero only the address portion while passing the flags through. Fixes: dbbdee94734b ("of/address: Merge all of the bus translation code") Cc: stable(a)vger.kernel.org Signed-off-by: Andrea della Porta <andrea.porta(a)suse.com> Tested-by: Herve Codina <herve.codina(a)bootlin.com> Link: https://lore.kernel.org/r/e51ae57874e58a9b349c35e2e877425ebc075d7a.17324418… Signed-off-by: Rob Herring (Arm) <robh(a)kernel.org> diff --git a/drivers/of/address.c b/drivers/of/address.c index c5b925ac469f..5b7ee3ed5296 100644 --- a/drivers/of/address.c +++ b/drivers/of/address.c @@ -459,7 +459,8 @@ static int of_translate_one(const struct device_node *parent, const struct of_bu } if (ranges == NULL || rlen == 0) { offset = of_read_number(addr, na); - memset(addr, 0, pna * 4); + /* set address to zero, pass flags through */ + memset(addr + pbus->flag_cells, 0, (pna - pbus->flag_cells) * 4); pr_debug("empty ranges; 1:1 translation\n"); goto finish; }

6 months, 1 week

1
0
0 0

FAILED: patch "[PATCH] of: address: Preserve the flags portion on 1:1 dma-ranges" failed to apply to 5.10-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 5.10-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-5.10.y git checkout FETCH_HEAD git cherry-pick -x 7f05e20b989ac33c9c0f8c2028ec0a566493548f # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2024122356-zen-badland-29c9@gregkh' --subject-prefix 'PATCH 5.10.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 7f05e20b989ac33c9c0f8c2028ec0a566493548f Mon Sep 17 00:00:00 2001 From: Andrea della Porta <andrea.porta(a)suse.com> Date: Sun, 24 Nov 2024 11:05:37 +0100 Subject: [PATCH] of: address: Preserve the flags portion on 1:1 dma-ranges mapping A missing or empty dma-ranges in a DT node implies a 1:1 mapping for dma translations. In this specific case, the current behaviour is to zero out the entire specifier so that the translation could be carried on as an offset from zero. This includes address specifier that has flags (e.g. PCI ranges). Once the flags portion has been zeroed, the translation chain is broken since the mapping functions will check the upcoming address specifier against mismatching flags, always failing the 1:1 mapping and its entire purpose of always succeeding. Set to zero only the address portion while passing the flags through. Fixes: dbbdee94734b ("of/address: Merge all of the bus translation code") Cc: stable(a)vger.kernel.org Signed-off-by: Andrea della Porta <andrea.porta(a)suse.com> Tested-by: Herve Codina <herve.codina(a)bootlin.com> Link: https://lore.kernel.org/r/e51ae57874e58a9b349c35e2e877425ebc075d7a.17324418… Signed-off-by: Rob Herring (Arm) <robh(a)kernel.org> diff --git a/drivers/of/address.c b/drivers/of/address.c index c5b925ac469f..5b7ee3ed5296 100644 --- a/drivers/of/address.c +++ b/drivers/of/address.c @@ -459,7 +459,8 @@ static int of_translate_one(const struct device_node *parent, const struct of_bu } if (ranges == NULL || rlen == 0) { offset = of_read_number(addr, na); - memset(addr, 0, pna * 4); + /* set address to zero, pass flags through */ + memset(addr + pbus->flag_cells, 0, (pna - pbus->flag_cells) * 4); pr_debug("empty ranges; 1:1 translation\n"); goto finish; }

6 months, 1 week

1
0
0 0

FAILED: patch "[PATCH] of: address: Preserve the flags portion on 1:1 dma-ranges" failed to apply to 5.4-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 5.4-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-5.4.y git checkout FETCH_HEAD git cherry-pick -x 7f05e20b989ac33c9c0f8c2028ec0a566493548f # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2024122347-fiscally-slacks-e686@gregkh' --subject-prefix 'PATCH 5.4.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 7f05e20b989ac33c9c0f8c2028ec0a566493548f Mon Sep 17 00:00:00 2001 From: Andrea della Porta <andrea.porta(a)suse.com> Date: Sun, 24 Nov 2024 11:05:37 +0100 Subject: [PATCH] of: address: Preserve the flags portion on 1:1 dma-ranges mapping A missing or empty dma-ranges in a DT node implies a 1:1 mapping for dma translations. In this specific case, the current behaviour is to zero out the entire specifier so that the translation could be carried on as an offset from zero. This includes address specifier that has flags (e.g. PCI ranges). Once the flags portion has been zeroed, the translation chain is broken since the mapping functions will check the upcoming address specifier against mismatching flags, always failing the 1:1 mapping and its entire purpose of always succeeding. Set to zero only the address portion while passing the flags through. Fixes: dbbdee94734b ("of/address: Merge all of the bus translation code") Cc: stable(a)vger.kernel.org Signed-off-by: Andrea della Porta <andrea.porta(a)suse.com> Tested-by: Herve Codina <herve.codina(a)bootlin.com> Link: https://lore.kernel.org/r/e51ae57874e58a9b349c35e2e877425ebc075d7a.17324418… Signed-off-by: Rob Herring (Arm) <robh(a)kernel.org> diff --git a/drivers/of/address.c b/drivers/of/address.c index c5b925ac469f..5b7ee3ed5296 100644 --- a/drivers/of/address.c +++ b/drivers/of/address.c @@ -459,7 +459,8 @@ static int of_translate_one(const struct device_node *parent, const struct of_bu } if (ranges == NULL || rlen == 0) { offset = of_read_number(addr, na); - memset(addr, 0, pna * 4); + /* set address to zero, pass flags through */ + memset(addr + pbus->flag_cells, 0, (pna - pbus->flag_cells) * 4); pr_debug("empty ranges; 1:1 translation\n"); goto finish; }

6 months, 1 week

1
0
0 0

[PATCH stable] udf: Fix directory iteration for longer tail extents

by Shreenidhi Shedi

Hello, Please consider applying the following patch to the 6.1.y stable tree. Commit: 1ea1cd11c72d1405a6b98440a9d5ea82dfa07166 Subject: udf: Fix directory iteration for longer tail extents Kernel Versions: 6.1.y Reason for inclusion: This patch resolves a performance issue that slows down directory traversal in UDF ISO images. The problem affects systems running kernel version 6.1.y. The fix has been merged into the mainline as commit 1ea1cd11c. No prep patches are needed. Link to upstream commit: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -- Shedi

6 months, 1 week

1
0
0 0

FAILED: patch "[PATCH] ceph: fix memory leak in ceph_direct_read_write()" failed to apply to 6.6-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 6.6-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.6.y git checkout FETCH_HEAD git cherry-pick -x 66e0c4f91461d17d48071695271c824620bed4ef # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2024122357-underuse-trousers-8f4d@gregkh' --subject-prefix 'PATCH 6.6.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 66e0c4f91461d17d48071695271c824620bed4ef Mon Sep 17 00:00:00 2001 From: Ilya Dryomov <idryomov(a)gmail.com> Date: Fri, 6 Dec 2024 17:32:59 +0100 Subject: [PATCH] ceph: fix memory leak in ceph_direct_read_write() The bvecs array which is allocated in iter_get_bvecs_alloc() is leaked and pages remain pinned if ceph_alloc_sparse_ext_map() fails. There is no need to delay the allocation of sparse_ext map until after the bvecs array is set up, so fix this by moving sparse_ext allocation a bit earlier. Also, make a similar adjustment in __ceph_sync_read() for consistency (a leak of the same kind in __ceph_sync_read() has been addressed differently). Cc: stable(a)vger.kernel.org Fixes: 03bc06c7b0bd ("ceph: add new mount option to enable sparse reads") Signed-off-by: Ilya Dryomov <idryomov(a)gmail.com> Reviewed-by: Alex Markuze <amarkuze(a)redhat.com> diff --git a/fs/ceph/file.c b/fs/ceph/file.c index 8e0400d461a2..67468d88f139 100644 --- a/fs/ceph/file.c +++ b/fs/ceph/file.c @@ -1116,6 +1116,16 @@ ssize_t __ceph_sync_read(struct inode *inode, loff_t *ki_pos, len = read_off + read_len - off; more = len < iov_iter_count(to); + op = &req->r_ops[0]; + if (sparse) { + extent_cnt = __ceph_sparse_read_ext_count(inode, read_len); + ret = ceph_alloc_sparse_ext_map(op, extent_cnt); + if (ret) { + ceph_osdc_put_request(req); + break; + } + } + num_pages = calc_pages_for(read_off, read_len); page_off = offset_in_page(off); pages = ceph_alloc_page_vector(num_pages, GFP_KERNEL); @@ -1129,16 +1139,6 @@ ssize_t __ceph_sync_read(struct inode *inode, loff_t *ki_pos, offset_in_page(read_off), false, true); - op = &req->r_ops[0]; - if (sparse) { - extent_cnt = __ceph_sparse_read_ext_count(inode, read_len); - ret = ceph_alloc_sparse_ext_map(op, extent_cnt); - if (ret) { - ceph_osdc_put_request(req); - break; - } - } - ceph_osdc_start_request(osdc, req); ret = ceph_osdc_wait_request(osdc, req); @@ -1551,6 +1551,16 @@ ceph_direct_read_write(struct kiocb *iocb, struct iov_iter *iter, break; } + op = &req->r_ops[0]; + if (sparse) { + extent_cnt = __ceph_sparse_read_ext_count(inode, size); + ret = ceph_alloc_sparse_ext_map(op, extent_cnt); + if (ret) { + ceph_osdc_put_request(req); + break; + } + } + len = iter_get_bvecs_alloc(iter, size, &bvecs, &num_pages); if (len < 0) { ceph_osdc_put_request(req); @@ -1560,6 +1570,8 @@ ceph_direct_read_write(struct kiocb *iocb, struct iov_iter *iter, if (len != size) osd_req_op_extent_update(req, 0, len); + osd_req_op_extent_osd_data_bvecs(req, 0, bvecs, num_pages, len); + /* * To simplify error handling, allow AIO when IO within i_size * or IO can be satisfied by single OSD request. @@ -1591,17 +1603,6 @@ ceph_direct_read_write(struct kiocb *iocb, struct iov_iter *iter, req->r_mtime = mtime; } - osd_req_op_extent_osd_data_bvecs(req, 0, bvecs, num_pages, len); - op = &req->r_ops[0]; - if (sparse) { - extent_cnt = __ceph_sparse_read_ext_count(inode, size); - ret = ceph_alloc_sparse_ext_map(op, extent_cnt); - if (ret) { - ceph_osdc_put_request(req); - break; - } - } - if (aio_req) { aio_req->total_len += len; aio_req->num_reqs++;

6 months, 1 week

1
0
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror December 2024