From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: stable@vger.kernel.org
Cc: "Greg Kroah-Hartman" <gregkh@linuxfoundation.org>,
patches@lists.linux.dev,
"Christian König" <christian.koenig@amd.com>,
"Sasha Levin" <sashal@kernel.org>,
"Tvrtko Ursulin" <tvrtko.ursulin@intel.com>
Subject: [PATCH 5.15 50/77] dma-buf: add dma_resv_for_each_fence_unlocked v8
Date: Tue, 15 Jul 2025 15:13:49 +0200 [thread overview]
Message-ID: <20250715130753.732374151@linuxfoundation.org> (raw)
In-Reply-To: <20250715130751.668489382@linuxfoundation.org>
5.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Christian König <christian.koenig@amd.com>
[ Upstream commit c921ff373b469ad7907cde219fa700909f59cac4 ]
Abstract the complexity of iterating over all the fences
in a dma_resv object.
The new loop handles the whole RCU and retry dance and
returns only fences where we can be sure we grabbed the
right one.
v2: fix accessing the shared fences while they might be freed,
improve kerneldoc, rename _cursor to _iter, add
dma_resv_iter_is_exclusive, add dma_resv_iter_begin/end
v3: restructor the code, move rcu_read_lock()/unlock() into the
iterator, add dma_resv_iter_is_restarted()
v4: fix NULL deref when no explicit fence exists, drop superflous
rcu_read_lock()/unlock() calls.
v5: fix typos in the documentation
v6: fix coding error when excl fence is NULL
v7: one more logic fix
v8: fix index check in dma_resv_iter_is_exclusive()
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> (v7)
Link: https://patchwork.freedesktop.org/patch/msgid/20211005113742.1101-2-christian.koenig@amd.com
Stable-dep-of: 2b95a7db6e0f ("dma-buf: fix timeout handling in dma_resv_wait_timeout v2")
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
drivers/dma-buf/dma-resv.c | 100 +++++++++++++++++++++++++++++++++++++
include/linux/dma-resv.h | 95 +++++++++++++++++++++++++++++++++++
2 files changed, 195 insertions(+)
diff --git a/drivers/dma-buf/dma-resv.c b/drivers/dma-buf/dma-resv.c
index e744fd87c63c8..c747b19c3be16 100644
--- a/drivers/dma-buf/dma-resv.c
+++ b/drivers/dma-buf/dma-resv.c
@@ -313,6 +313,106 @@ void dma_resv_add_excl_fence(struct dma_resv *obj, struct dma_fence *fence)
}
EXPORT_SYMBOL(dma_resv_add_excl_fence);
+/**
+ * dma_resv_iter_restart_unlocked - restart the unlocked iterator
+ * @cursor: The dma_resv_iter object to restart
+ *
+ * Restart the unlocked iteration by initializing the cursor object.
+ */
+static void dma_resv_iter_restart_unlocked(struct dma_resv_iter *cursor)
+{
+ cursor->seq = read_seqcount_begin(&cursor->obj->seq);
+ cursor->index = -1;
+ if (cursor->all_fences)
+ cursor->fences = dma_resv_shared_list(cursor->obj);
+ else
+ cursor->fences = NULL;
+ cursor->is_restarted = true;
+}
+
+/**
+ * dma_resv_iter_walk_unlocked - walk over fences in a dma_resv obj
+ * @cursor: cursor to record the current position
+ *
+ * Return all the fences in the dma_resv object which are not yet signaled.
+ * The returned fence has an extra local reference so will stay alive.
+ * If a concurrent modify is detected the whole iteration is started over again.
+ */
+static void dma_resv_iter_walk_unlocked(struct dma_resv_iter *cursor)
+{
+ struct dma_resv *obj = cursor->obj;
+
+ do {
+ /* Drop the reference from the previous round */
+ dma_fence_put(cursor->fence);
+
+ if (cursor->index == -1) {
+ cursor->fence = dma_resv_excl_fence(obj);
+ cursor->index++;
+ if (!cursor->fence)
+ continue;
+
+ } else if (!cursor->fences ||
+ cursor->index >= cursor->fences->shared_count) {
+ cursor->fence = NULL;
+ break;
+
+ } else {
+ struct dma_resv_list *fences = cursor->fences;
+ unsigned int idx = cursor->index++;
+
+ cursor->fence = rcu_dereference(fences->shared[idx]);
+ }
+ cursor->fence = dma_fence_get_rcu(cursor->fence);
+ if (!cursor->fence || !dma_fence_is_signaled(cursor->fence))
+ break;
+ } while (true);
+}
+
+/**
+ * dma_resv_iter_first_unlocked - first fence in an unlocked dma_resv obj.
+ * @cursor: the cursor with the current position
+ *
+ * Returns the first fence from an unlocked dma_resv obj.
+ */
+struct dma_fence *dma_resv_iter_first_unlocked(struct dma_resv_iter *cursor)
+{
+ rcu_read_lock();
+ do {
+ dma_resv_iter_restart_unlocked(cursor);
+ dma_resv_iter_walk_unlocked(cursor);
+ } while (read_seqcount_retry(&cursor->obj->seq, cursor->seq));
+ rcu_read_unlock();
+
+ return cursor->fence;
+}
+EXPORT_SYMBOL(dma_resv_iter_first_unlocked);
+
+/**
+ * dma_resv_iter_next_unlocked - next fence in an unlocked dma_resv obj.
+ * @cursor: the cursor with the current position
+ *
+ * Returns the next fence from an unlocked dma_resv obj.
+ */
+struct dma_fence *dma_resv_iter_next_unlocked(struct dma_resv_iter *cursor)
+{
+ bool restart;
+
+ rcu_read_lock();
+ cursor->is_restarted = false;
+ restart = read_seqcount_retry(&cursor->obj->seq, cursor->seq);
+ do {
+ if (restart)
+ dma_resv_iter_restart_unlocked(cursor);
+ dma_resv_iter_walk_unlocked(cursor);
+ restart = true;
+ } while (read_seqcount_retry(&cursor->obj->seq, cursor->seq));
+ rcu_read_unlock();
+
+ return cursor->fence;
+}
+EXPORT_SYMBOL(dma_resv_iter_next_unlocked);
+
/**
* dma_resv_copy_fences - Copy all fences from src to dst.
* @dst: the destination reservation object
diff --git a/include/linux/dma-resv.h b/include/linux/dma-resv.h
index e1ca2080a1ff1..ef9f94ce2cc38 100644
--- a/include/linux/dma-resv.h
+++ b/include/linux/dma-resv.h
@@ -75,6 +75,101 @@ struct dma_resv {
struct dma_resv_list __rcu *fence;
};
+/**
+ * struct dma_resv_iter - current position into the dma_resv fences
+ *
+ * Don't touch this directly in the driver, use the accessor function instead.
+ */
+struct dma_resv_iter {
+ /** @obj: The dma_resv object we iterate over */
+ struct dma_resv *obj;
+
+ /** @all_fences: If all fences should be returned */
+ bool all_fences;
+
+ /** @fence: the currently handled fence */
+ struct dma_fence *fence;
+
+ /** @seq: sequence number to check for modifications */
+ unsigned int seq;
+
+ /** @index: index into the shared fences */
+ unsigned int index;
+
+ /** @fences: the shared fences */
+ struct dma_resv_list *fences;
+
+ /** @is_restarted: true if this is the first returned fence */
+ bool is_restarted;
+};
+
+struct dma_fence *dma_resv_iter_first_unlocked(struct dma_resv_iter *cursor);
+struct dma_fence *dma_resv_iter_next_unlocked(struct dma_resv_iter *cursor);
+
+/**
+ * dma_resv_iter_begin - initialize a dma_resv_iter object
+ * @cursor: The dma_resv_iter object to initialize
+ * @obj: The dma_resv object which we want to iterate over
+ * @all_fences: If all fences should be returned or just the exclusive one
+ */
+static inline void dma_resv_iter_begin(struct dma_resv_iter *cursor,
+ struct dma_resv *obj,
+ bool all_fences)
+{
+ cursor->obj = obj;
+ cursor->all_fences = all_fences;
+ cursor->fence = NULL;
+}
+
+/**
+ * dma_resv_iter_end - cleanup a dma_resv_iter object
+ * @cursor: the dma_resv_iter object which should be cleaned up
+ *
+ * Make sure that the reference to the fence in the cursor is properly
+ * dropped.
+ */
+static inline void dma_resv_iter_end(struct dma_resv_iter *cursor)
+{
+ dma_fence_put(cursor->fence);
+}
+
+/**
+ * dma_resv_iter_is_exclusive - test if the current fence is the exclusive one
+ * @cursor: the cursor of the current position
+ *
+ * Returns true if the currently returned fence is the exclusive one.
+ */
+static inline bool dma_resv_iter_is_exclusive(struct dma_resv_iter *cursor)
+{
+ return cursor->index == 0;
+}
+
+/**
+ * dma_resv_iter_is_restarted - test if this is the first fence after a restart
+ * @cursor: the cursor with the current position
+ *
+ * Return true if this is the first fence in an iteration after a restart.
+ */
+static inline bool dma_resv_iter_is_restarted(struct dma_resv_iter *cursor)
+{
+ return cursor->is_restarted;
+}
+
+/**
+ * dma_resv_for_each_fence_unlocked - unlocked fence iterator
+ * @cursor: a struct dma_resv_iter pointer
+ * @fence: the current fence
+ *
+ * Iterate over the fences in a struct dma_resv object without holding the
+ * &dma_resv.lock and using RCU instead. The cursor needs to be initialized
+ * with dma_resv_iter_begin() and cleaned up with dma_resv_iter_end(). Inside
+ * the iterator a reference to the dma_fence is held and the RCU lock dropped.
+ * When the dma_resv is modified the iteration starts over again.
+ */
+#define dma_resv_for_each_fence_unlocked(cursor, fence) \
+ for (fence = dma_resv_iter_first_unlocked(cursor); \
+ fence; fence = dma_resv_iter_next_unlocked(cursor))
+
#define dma_resv_held(obj) lockdep_is_held(&(obj)->lock.base)
#define dma_resv_assert_held(obj) lockdep_assert_held(&(obj)->lock.base)
--
2.39.5
next prev parent reply other threads:[~2025-07-15 13:29 UTC|newest]
Thread overview: 86+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-07-15 13:12 [PATCH 5.15 00/77] 5.15.189-rc1 review Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 01/77] drm/exynos: exynos7_drm_decon: add vblank check in IRQ handling Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 02/77] ASoC: fsl_asrc: use internal measured ratio for non-ideal ratio mode Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 03/77] perf: Revert to requiring CAP_SYS_ADMIN for uprobes Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 04/77] fix proc_sys_compare() handling of in-lookup dentries Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 05/77] netlink: Fix wraparounds of sk->sk_rmem_alloc Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 06/77] tipc: Fix use-after-free in tipc_conn_close() Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 07/77] vsock: Fix transport_{g2h,h2g} TOCTOU Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 08/77] vsock: Fix transport_* TOCTOU Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 09/77] vsock: Fix IOCTL_VM_SOCKETS_GET_LOCAL_CID to check also `transport_local` Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 10/77] net: phy: smsc: Fix Auto-MDIX configuration when disabled by strap Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 11/77] net: phy: smsc: Fix link failure in forced mode with Auto-MDIX Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 12/77] atm: clip: Fix potential null-ptr-deref in to_atmarpd() Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 13/77] atm: clip: Fix memory leak of struct clip_vcc Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 14/77] atm: clip: Fix infinite recursive call of clip_push() Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 15/77] atm: clip: Fix NULL pointer dereference in vcc_sendmsg() Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 16/77] net/sched: Abort __tc_modify_qdisc if parent class does not exist Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 17/77] fs/proc: do_task_stat: use __for_each_thread() Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 18/77] ice: safer stats processing Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 19/77] rxrpc: Fix oops due to non-existence of prealloc backlog struct Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 20/77] bpf: fix precision backtracking instruction iteration Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 21/77] thermal/int340x_thermal: handle data_vault when the value is ZERO_SIZE_PTR Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 22/77] aoe: avoid potential deadlock at set_capacity Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 23/77] bpf, sockmap: Fix skb refcnt race after locking changes Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 24/77] jfs: fix null ptr deref in dtInsertEntry Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 25/77] xen: replace xen_remap() with memremap() Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 26/77] x86/mce/amd: Fix threshold limit reset Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 27/77] x86/mce: Dont remove sysfs if thresholding sysfs init fails Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 28/77] x86/mce: Make sure CMCI banks are cleared during shutdown on Intel Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 29/77] gre: Fix IPv6 multicast route creation Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 30/77] pinctrl: qcom: msm: mark certain pins as invalid for interrupts Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 31/77] drm/sched: Increment job count before swapping tail spsc queue Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 32/77] drm/gem: Fix race in drm_gem_handle_create_tail() Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 33/77] usb: gadget: u_serial: Fix race condition in TTY wakeup Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 34/77] Revert "ACPI: battery: negate current when discharging" Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 35/77] btrfs: propagate last_unlink_trans earlier when doing a rmdir Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 36/77] btrfs: use btrfs_record_snapshot_destroy() during rmdir Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 37/77] RDMA/mlx5: Fix vport loopback for MPV device Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 38/77] pwm: mediatek: Ensure to disable clocks in error path Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 39/77] netlink: Fix rmem check in netlink_broadcast_deliver() Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 40/77] netlink: make sure we allow at least one dump skb Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 41/77] virtio-net: ensure the received length does not exceed allocated size Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 42/77] xhci: Allow RPM on the USB controller (1022:43f7) by default Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 43/77] usb: xhci: quirk for data loss in ISOC transfers Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 44/77] xhci: Disable stream for xHC controller with XHCI_BROKEN_STREAMS Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 45/77] Input: xpad - support Acer NGR 200 Controller Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 46/77] usb:cdnsp: remove TRB_FLUSH_ENDPOINT command Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 47/77] usb: cdnsp: Replace snprintf() with the safer scnprintf() variant Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 48/77] usb: cdnsp: Fix issue with CV Bad Descriptor test Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 49/77] usb: dwc3: Abort suspend on soft disconnect failure Greg Kroah-Hartman
2025-07-15 13:13 ` Greg Kroah-Hartman [this message]
2025-07-15 13:13 ` [PATCH 5.15 51/77] dma-buf: use new iterator in dma_resv_wait_timeout Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 52/77] dma-buf: fix timeout handling in dma_resv_wait_timeout v2 Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 53/77] platform/x86: think-lmi: Fix sysfs group cleanup Greg Kroah-Hartman
2025-07-15 18:38 ` Kurt Borja
2025-07-15 19:00 ` Kurt Borja
2025-07-15 19:04 ` Sasha Levin
2025-07-15 19:22 ` Kurt Borja
2025-07-15 19:12 ` Sasha Levin
2025-07-15 21:18 ` Kurt Borja
2025-07-15 13:13 ` [PATCH 5.15 54/77] wifi: zd1211rw: Fix potential NULL pointer dereference in zd_mac_tx_to_dev() Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 55/77] md/raid1: Fix stack memory use after return in raid1_reshape Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 56/77] raid10: cleanup memleak at raid10_make_request Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 57/77] nbd: fix uaf in nbd_genl_connect() error path Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 58/77] smb: server: make use of rdma_destroy_qp() Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 59/77] ksmbd: fix a mount write count leak in ksmbd_vfs_kern_path_locked() Greg Kroah-Hartman
2025-07-15 13:13 ` [PATCH 5.15 60/77] netfilter: flowtable: account for Ethernet header in nf_flow_pppoe_proto() Greg Kroah-Hartman
2025-07-15 13:14 ` [PATCH 5.15 61/77] net: appletalk: Fix device refcount leak in atrtr_create() Greg Kroah-Hartman
2025-07-15 13:14 ` [PATCH 5.15 62/77] net: phy: microchip: limit 100M workaround to link-down events on LAN88xx Greg Kroah-Hartman
2025-07-15 13:14 ` [PATCH 5.15 63/77] can: m_can: m_can_handle_lost_msg(): downgrade msg lost in rx message to debug level Greg Kroah-Hartman
2025-07-15 13:14 ` [PATCH 5.15 64/77] net: ll_temac: Fix missing tx_pending check in ethtools_set_ringparam() Greg Kroah-Hartman
2025-07-15 13:14 ` [PATCH 5.15 65/77] bnxt_en: Fix DCB ETS validation Greg Kroah-Hartman
2025-07-15 13:14 ` [PATCH 5.15 66/77] bnxt_en: Set DMA unmap len correctly for XDP_REDIRECT Greg Kroah-Hartman
2025-07-15 13:14 ` [PATCH 5.15 67/77] atm: idt77252: Add missing `dma_map_error()` Greg Kroah-Hartman
2025-07-15 13:14 ` [PATCH 5.15 68/77] um: vector: Reduce stack usage in vector_eth_configure() Greg Kroah-Hartman
2025-07-15 13:14 ` [PATCH 5.15 69/77] net: usb: qmi_wwan: add SIMCom 8230C composition Greg Kroah-Hartman
2025-07-15 13:14 ` [PATCH 5.15 70/77] HID: lenovo: Add support for ThinkPad X1 Tablet Thin Keyboard Gen2 Greg Kroah-Hartman
2025-07-15 13:14 ` [PATCH 5.15 71/77] vt: add missing notification when switching back to text mode Greg Kroah-Hartman
2025-07-15 13:14 ` [PATCH 5.15 72/77] HID: Add IGNORE quirk for SMARTLINKTECHNOLOGY Greg Kroah-Hartman
2025-07-15 13:14 ` [PATCH 5.15 73/77] HID: quirks: Add quirk for 2 Chicony Electronics HP 5MP Cameras Greg Kroah-Hartman
2025-07-15 13:14 ` [PATCH 5.15 74/77] Input: atkbd - do not skip atkbd_deactivate() when skipping ATKBD_CMD_GETID Greg Kroah-Hartman
2025-07-15 13:14 ` [PATCH 5.15 75/77] vhost-scsi: protect vq->log_used with vq->mutex Greg Kroah-Hartman
2025-07-15 13:14 ` [PATCH 5.15 76/77] x86/mm: Disable hugetlb page table sharing on 32-bit Greg Kroah-Hartman
2025-07-15 13:14 ` [PATCH 5.15 77/77] x86: Fix X86_FEATURE_VERW_CLEAR definition Greg Kroah-Hartman
2025-07-15 16:20 ` [PATCH 5.15 00/77] 5.15.189-rc1 review Vijayendra Suman
2025-07-16 14:59 ` Shuah Khan
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20250715130753.732374151@linuxfoundation.org \
--to=gregkh@linuxfoundation.org \
--cc=christian.koenig@amd.com \
--cc=patches@lists.linux.dev \
--cc=sashal@kernel.org \
--cc=stable@vger.kernel.org \
--cc=tvrtko.ursulin@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.