mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Sasha Levin <Alexander.Levin@microsoft.com>
To: "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	"stable@vger.kernel.org" <stable@vger.kernel.org>
Cc: "Chris Wilson" <chris@chris-wilson.co.uk>,
	"Rob Clark" <robdclark@gmail.com>,
	"Gustavo Padovan" <gustavo.padovan@collabora.co.uk>,
	"Sumit Semwal" <sumit.semwal@linaro.org>,
	"Christian König" <christian.koenig@amd.com>,
	"Sasha Levin" <Alexander.Levin@microsoft.com>
Subject: [PATCH AUTOSEL for 4.14 02/84] dma-buf/fence: Fix lock inversion within dma-fence-array
Date: Sat, 3 Mar 2018 22:26:01 +0000	[thread overview]
Message-ID: <20180303222518.26271-2-alexander.levin@microsoft.com> (raw)
In-Reply-To: <20180303222518.26271-1-alexander.levin@microsoft.com>

From: Chris Wilson <chris@chris-wilson.co.uk>

[ Upstream commit 03e4e0a9e02cf703da331ff6cfd57d0be9bf5692 ]

Ages ago Rob Clark noted,

"Currently with fence-array, we have a potential deadlock situation.  If
we fence_add_callback() on an array-fence, the array-fence's lock is
acquired first, and in it's ->enable_signaling() callback, it will install
cbs on it's array-member fences, so the array-member's lock is acquired
second.

But in the signal path, the array-member's lock is acquired first, and
the array-fence's lock acquired second."

Rob proposed either extensive changes to dma-fence to unnest the
fence-array signaling, or to defer the signaling onto a workqueue. This
is a more refined version of the later, that should keep the latency
of the fence signaling to a minimum by using an irq-work, which is
executed asap.

Reported-by: Rob Clark <robdclark@gmail.com>
Suggested-by: Rob Clark <robdclark@gmail.com>
References: 1476635975-21981-1-git-send-email-robdclark@gmail.com
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Rob Clark <robdclark@gmail.com>
Cc: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Cc: Sumit Semwal <sumit.semwal@linaro.org>
Cc: Christian König <christian.koenig@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Sumit Semwal <sumit.semwal@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20171114162719.30958-1-chris@chris-wilson.co.uk
Signed-off-by: Sasha Levin <alexander.levin@microsoft.com>
---
 drivers/base/Kconfig              |  1 +
 drivers/dma-buf/dma-fence-array.c | 14 ++++++++++++--
 include/linux/dma-fence-array.h   |  3 +++
 3 files changed, 16 insertions(+), 2 deletions(-)

diff --git a/drivers/base/Kconfig b/drivers/base/Kconfig
index 2415ad9f6dd4..49fd50fccd48 100644
--- a/drivers/base/Kconfig
+++ b/drivers/base/Kconfig
@@ -249,6 +249,7 @@ config DMA_SHARED_BUFFER
 	bool
 	default n
 	select ANON_INODES
+	select IRQ_WORK
 	help
 	  This option enables the framework for buffer-sharing between
 	  multiple drivers. A buffer is associated with a file using driver
diff --git a/drivers/dma-buf/dma-fence-array.c b/drivers/dma-buf/dma-fence-array.c
index 0350829ba62e..dd1edfb27b61 100644
--- a/drivers/dma-buf/dma-fence-array.c
+++ b/drivers/dma-buf/dma-fence-array.c
@@ -31,6 +31,14 @@ static const char *dma_fence_array_get_timeline_name(struct dma_fence *fence)
 	return "unbound";
 }
 
+static void irq_dma_fence_array_work(struct irq_work *wrk)
+{
+	struct dma_fence_array *array = container_of(wrk, typeof(*array), work);
+
+	dma_fence_signal(&array->base);
+	dma_fence_put(&array->base);
+}
+
 static void dma_fence_array_cb_func(struct dma_fence *f,
 				    struct dma_fence_cb *cb)
 {
@@ -39,8 +47,9 @@ static void dma_fence_array_cb_func(struct dma_fence *f,
 	struct dma_fence_array *array = array_cb->array;
 
 	if (atomic_dec_and_test(&array->num_pending))
-		dma_fence_signal(&array->base);
-	dma_fence_put(&array->base);
+		irq_work_queue(&array->work);
+	else
+		dma_fence_put(&array->base);
 }
 
 static bool dma_fence_array_enable_signaling(struct dma_fence *fence)
@@ -136,6 +145,7 @@ struct dma_fence_array *dma_fence_array_create(int num_fences,
 	spin_lock_init(&array->lock);
 	dma_fence_init(&array->base, &dma_fence_array_ops, &array->lock,
 		       context, seqno);
+	init_irq_work(&array->work, irq_dma_fence_array_work);
 
 	array->num_fences = num_fences;
 	atomic_set(&array->num_pending, signal_on_any ? 1 : num_fences);
diff --git a/include/linux/dma-fence-array.h b/include/linux/dma-fence-array.h
index 332a5420243c..bc8940ca280d 100644
--- a/include/linux/dma-fence-array.h
+++ b/include/linux/dma-fence-array.h
@@ -21,6 +21,7 @@
 #define __LINUX_DMA_FENCE_ARRAY_H
 
 #include <linux/dma-fence.h>
+#include <linux/irq_work.h>
 
 /**
  * struct dma_fence_array_cb - callback helper for fence array
@@ -47,6 +48,8 @@ struct dma_fence_array {
 	unsigned num_fences;
 	atomic_t num_pending;
 	struct dma_fence **fences;
+
+	struct irq_work work;
 };
 
 extern const struct dma_fence_ops dma_fence_array_ops;
-- 
2.14.1

  reply	other threads:[~2018-03-03 22:28 UTC|newest]

Thread overview: 84+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-03-03 22:26 [PATCH AUTOSEL for 4.14 01/84] drm/edid: set ELD connector type in drm_edid_to_eld() Sasha Levin
2018-03-03 22:26 ` Sasha Levin [this message]
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 03/84] video/hdmi: Allow "empty" HDMI infoframes Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 04/84] HID: multitouch: Only look at non touch fields in first packet of a frame Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 05/84] HID: elo: clear BTN_LEFT mapping Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 06/84] iwlwifi: mvm: rs: don't override the rate history in the search cycle Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 07/84] ARM: dts: koelsch: Move cec_clock to root node Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 08/84] clk: meson: gxbb: fix wrong clock for SARADC/SANA Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 09/84] ARM: dts: exynos: Correct Trats2 panel reset line Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 10/84] drm/amdgpu: fix get_max_engine_clock_in_mhz Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 12/84] typec: tcpm: fusb302: Resolve out of order messaging events Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 11/84] staging: rtl8822be: fix missing null check on dev_alloc_skb return Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 13/84] USB: ledtrig-usbport: fix of-node leak Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 15/84] sched: Stop resched_cpu() from sending IPIs to offline CPUs Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 14/84] sched: Stop switched_to_rt() " Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 17/84] crypto: cavium - fix memory leak on info Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 16/84] crypto: ecc - Fix NULL pointer deref. on no default_rng Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 18/84] test_firmware: fix setting old custom fw path back on exit Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 19/84] ASoC: fsl_ssi: only enable proper channel slots in AC'97 mode Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 20/84] net: ieee802154: adf7242: Fix bug if defined DEBUG Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 21/84] rtc: brcmstb-waketimer: fix error handling in brcmstb_waketmr_probe() Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 22/84] net: xfrm: allow clearing socket xfrm policies Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 23/84] mtd: nand: fix interpretation of NAND_CMD_NONE in nand_command[_lp]() Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 24/84] net: thunderx: Set max queue count taking XDP_TX into account Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 25/84] ARM: dts: am335x-pepper: Fix the audio CODEC's reset pin Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 27/84] mtd: nand: ifc: update bufnum mask for ver >= 2.0.0 Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 26/84] ARM: dts: omap3-n900: Fix the audio CODEC's reset pin Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 29/84] xfrm: Fix xfrm_replay_overflow_offload_esn Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 28/84] userns: Don't fail follow_automount based on s_user_ns Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 30/84] leds: pm8058: Silence pointer to integer size warning Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 31/84] clk: ti: clkctrl: add support for retrying failed init Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 32/84] ASoC: tlv320aic31xx: Handle inverted BCLK in non-DSP modes Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 33/84] power: supply: ab8500_charger: Fix an error handling path Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 34/84] power: supply: ab8500_charger: Bail out in case of error in 'ab8500_charger_init_hw_registers()' Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 35/84] drm/etnaviv: make THERMAL selectable Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 36/84] iio: adc: ina2xx: Shift bus voltage register to mask flag bits Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 38/84] ath10k: update tdls teardown state to target Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 37/84] iio: health: max30102: Add power enable parameter to get_temp function Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 40/84] dmaengine: bcm2835-dma: Use vchan_terminate_vdesc() instead of desc_free Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 39/84] cpufreq: Fix governor module removal race Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 41/84] dmaengine: amba-pl08x: Use vchan_terminate_vdesc() instead of desc_free Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 43/84] scsi: ses: don't ask for diagnostic pages repeatedly during probe Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 42/84] drm/amdgpu:fix random missing of FLR NOTIFY Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 44/84] pwm: stmpe: Fix wrong register offset for hwpwm=2 case Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 45/84] drm/sun4i: Fix format mask in DE2 driver Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 46/84] pinctrl: sh-pfc: r8a7791: Add can_clk function Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 47/84] pinctrl: sh-pfc: r8a7795-es1: Fix MOD_SEL1 bit[25:24] to 0x3 when using STP_ISEN_1_D Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 48/84] perf annotate: Fix unnecessary memory allocation for s390x Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 50/84] iwlwifi: mvm: avoid dumping assert log when device is stopped Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 49/84] perf annotate: Fix objdump comment parsing for Intel mov dissassembly Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 51/84] drm/amdgpu:fix virtual dce bug Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 52/84] clk: qcom: msm8916: fix mnd_width for codec_digcodec Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 53/84] mwifiex: cfg80211: do not change virtual interface during scan processing Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 54/84] ath10k: fix invalid STS_CAP_OFFSET_MASK Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 56/84] spi: sun6i: disable/unprepare clocks on remove Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 55/84] tools/usbip: fixes build with musl libc toolchain Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 58/84] scsi: core: scsi_get_device_flags_keyed(): Always return device flags Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 57/84] bnxt_en: Don't print "Link speed -1 no longer supported" messages Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 59/84] scsi: devinfo: apply to HP XP the same flags as Hitachi VSP Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 60/84] scsi: dh: add new rdac devices Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 61/84] media: vsp1: Prevent suspending and resuming DRM pipelines Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 62/84] dm raid: fix raid set size revalidation Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 64/84] media: davinci: vpif_capture: add NULL check on devm_kzalloc return value Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 63/84] media: cpia2: Fix a couple off by one bugs Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 65/84] virtio_net: Disable interrupts if napi_complete_done rescheduled napi Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 66/84] net: sched: drop qdisc_reset from dev_graft_qdisc Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 67/84] veth: set peer GSO values Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 69/84] powerpc/modules: Don't try to restore r2 after a sibling call Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 68/84] drm/amdkfd: Fix memory leaks in kfd topology Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 70/84] powerpc/64: Don't trace irqs-off at interrupt return to soft-disabled context Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 71/84] arm64: dts: renesas: salvator-common: Add EthernetAVB PHY reset Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 72/84] agp/intel: Flush all chipset writes after updating the GGTT Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 73/84] mac80211_hwsim: enforce PS_MANUAL_POLL to be set after PS_ENABLED Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 74/84] mac80211: remove BUG() when interface type is invalid Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 75/84] crypto: caam/qi - use correct print specifier for size_t Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 76/84] ASoC: nuc900: Fix a loop timeout test Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 78/84] Fix misannotated out-of-line _copy_to_user() Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 77/84] mmc: mmc_test: Ensure command queue is disabled for testing Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 79/84] ipvlan: add L2 check for packets arriving via virtual devices Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 80/84] rcutorture/configinit: Fix build directory error message Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 81/84] locking/locktorture: Fix num reader/writer corner cases Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 82/84] ima: relax requiring a file signature for new files with zero length Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 83/84] IB/mlx5: revisit -Wmaybe-uninitialized warning Sasha Levin
2018-03-03 22:26 ` [PATCH AUTOSEL for 4.14 84/84] dmaengine: qcom_hidma: check pending interrupts Sasha Levin

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=20180303222518.26271-2-alexander.levin@microsoft.com \
    --to=alexander.levin@microsoft.com \
    --cc=chris@chris-wilson.co.uk \
    --cc=christian.koenig@amd.com \
    --cc=gustavo.padovan@collabora.co.uk \
    --cc=linux-kernel@vger.kernel.org \
    --cc=robdclark@gmail.com \
    --cc=stable@vger.kernel.org \
    --cc=sumit.semwal@linaro.org \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®