From: Sasha Levin <Alexander.Levin@microsoft.com>
To: "stable@vger.kernel.org" <stable@vger.kernel.org>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Cc: Ben Skeggs <bskeggs@redhat.com>,
Sasha Levin <Alexander.Levin@microsoft.com>
Subject: [PATCH AUTOSEL 4.14 19/37] drm/nouveau/disp: fix DP disable race
Date: Mon, 1 Oct 2018 00:39:05 +0000 [thread overview]
Message-ID: <20181001003850.147107-19-alexander.levin@microsoft.com> (raw)
In-Reply-To: <20181001003850.147107-1-alexander.levin@microsoft.com>
From: Ben Skeggs <bskeggs@redhat.com>
[ Upstream commit e04cfdc9b7398c60dbc70212415ea63b6c6a93ae ]
If a HPD pulse signalling the need to retrain the link occurs between
the KMS driver releasing the output and the supervisor interrupt that
finishes the teardown, it was possible get a NULL-ptr deref.
Avoid this by marking the link as inactive earlier.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Sasha Levin <alexander.levin@microsoft.com>
---
drivers/gpu/drm/nouveau/nvkm/engine/disp/dp.c | 17 ++++++++++++-----
drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c | 6 +++---
drivers/gpu/drm/nouveau/nvkm/engine/disp/outp.c | 2 ++
drivers/gpu/drm/nouveau/nvkm/engine/disp/outp.h | 3 ++-
4 files changed, 19 insertions(+), 9 deletions(-)
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dp.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dp.c
index 7c5bed29ffef..6160a6158cf2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dp.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dp.c
@@ -412,14 +412,10 @@ nvkm_dp_train(struct nvkm_dp *dp, u32 dataKBps)
}
static void
-nvkm_dp_release(struct nvkm_outp *outp, struct nvkm_ior *ior)
+nvkm_dp_disable(struct nvkm_outp *outp, struct nvkm_ior *ior)
{
struct nvkm_dp *dp = nvkm_dp(outp);
- /* Prevent link from being retrained if sink sends an IRQ. */
- atomic_set(&dp->lt.done, 0);
- ior->dp.nr = 0;
-
/* Execute DisableLT script from DP Info Table. */
nvbios_init(&ior->disp->engine.subdev, dp->info.script[4],
init.outp = &dp->outp.info;
@@ -428,6 +424,16 @@ nvkm_dp_release(struct nvkm_outp *outp, struct nvkm_ior *ior)
);
}
+static void
+nvkm_dp_release(struct nvkm_outp *outp)
+{
+ struct nvkm_dp *dp = nvkm_dp(outp);
+
+ /* Prevent link from being retrained if sink sends an IRQ. */
+ atomic_set(&dp->lt.done, 0);
+ dp->outp.ior->dp.nr = 0;
+}
+
static int
nvkm_dp_acquire(struct nvkm_outp *outp)
{
@@ -576,6 +582,7 @@ nvkm_dp_func = {
.fini = nvkm_dp_fini,
.acquire = nvkm_dp_acquire,
.release = nvkm_dp_release,
+ .disable = nvkm_dp_disable,
};
static int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
index 0c570dbd3021..bc18a96bc61a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
@@ -436,11 +436,11 @@ nv50_disp_super_2_0(struct nv50_disp *disp, struct nvkm_head *head)
nv50_disp_super_ied_off(head, ior, 2);
/* If we're shutting down the OR's only active head, execute
- * the output path's release function.
+ * the output path's disable function.
*/
if (ior->arm.head == (1 << head->id)) {
- if ((outp = ior->arm.outp) && outp->func->release)
- outp->func->release(outp, ior);
+ if ((outp = ior->arm.outp) && outp->func->disable)
+ outp->func->disable(outp, ior);
}
}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/outp.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/outp.c
index be9e7f8c3b23..bbba77ff9385 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/outp.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/outp.c
@@ -93,6 +93,8 @@ nvkm_outp_release(struct nvkm_outp *outp, u8 user)
if (ior) {
outp->acquired &= ~user;
if (!outp->acquired) {
+ if (outp->func->release && outp->ior)
+ outp->func->release(outp);
outp->ior->asy.outp = NULL;
outp->ior = NULL;
}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/outp.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/outp.h
index ea84d7d5741a..97196f802924 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/outp.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/outp.h
@@ -41,7 +41,8 @@ struct nvkm_outp_func {
void (*init)(struct nvkm_outp *);
void (*fini)(struct nvkm_outp *);
int (*acquire)(struct nvkm_outp *);
- void (*release)(struct nvkm_outp *, struct nvkm_ior *);
+ void (*release)(struct nvkm_outp *);
+ void (*disable)(struct nvkm_outp *, struct nvkm_ior *);
};
#define OUTP_MSG(o,l,f,a...) do { \
--
2.17.1
next prev parent reply other threads:[~2018-10-01 0:48 UTC|newest]
Thread overview: 37+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-10-01 0:38 [PATCH AUTOSEL 4.14 01/37] netfilter: xt_cluster: add dependency on conntrack module Sasha Levin
2018-10-01 0:38 ` [PATCH AUTOSEL 4.14 02/37] HID: add support for Apple Magic Keyboards Sasha Levin
2018-10-01 0:38 ` [PATCH AUTOSEL 4.14 03/37] usb: gadget: fotg210-udc: Fix memory leak of fotg210->ep[i] Sasha Levin
2018-10-01 0:38 ` [PATCH AUTOSEL 4.14 05/37] HID: hid-saitek: Add device ID for RAT 7 Contagion Sasha Levin
2018-10-01 0:38 ` [PATCH AUTOSEL 4.14 04/37] pinctrl: msm: Really mask level interrupts to prevent latching Sasha Levin
2018-10-01 0:38 ` [PATCH AUTOSEL 4.14 06/37] scsi: iscsi: target: Set conn->sess to NULL when iscsi_login_set_conn_values fails Sasha Levin
2018-10-01 0:38 ` [PATCH AUTOSEL 4.14 07/37] scsi: qedi: Add the CRC size within iSCSI NVM image Sasha Levin
2018-10-01 0:38 ` [PATCH AUTOSEL 4.14 08/37] perf evsel: Fix potential null pointer dereference in perf_evsel__new_idx() Sasha Levin
2018-10-01 0:38 ` [PATCH AUTOSEL 4.14 09/37] perf util: Fix bad memory access in trace info Sasha Levin
2018-10-01 0:38 ` [PATCH AUTOSEL 4.14 10/37] perf probe powerpc: Ignore SyS symbols irrespective of endianness Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 11/37] netfilter: nf_tables: release chain in flushing set Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 12/37] Revert "iio: temperature: maxim_thermocouple: add MAX31856 part" Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 13/37] RDMA/ucma: check fd type in ucma_migrate_id() Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 14/37] HID: sensor-hub: Restore fixup for Lenovo ThinkPad Helix 2 sensor hub report Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 15/37] USB: yurex: Check for truncation in yurex_read() Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 16/37] nvmet-rdma: fix possible bogus dereference under heavy load Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 18/37] drm/nouveau/TBDdevinit: don't fail when PMU/PRE_OS is missing from VBIOS Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 17/37] net/mlx5: Consider PCI domain in search for next dev Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 20/37] dm raid: fix rebuild of specific devices by updating superblock Sasha Levin
2018-10-01 0:39 ` Sasha Levin [this message]
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 21/37] fs/cifs: suppress a string overflow warning Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 22/37] net: ena: fix driver when PAGE_SIZE == 64kB Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 23/37] net: ena: fix missing calls to READ_ONCE Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 25/37] dm thin metadata: try to avoid ever aborting transactions Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 24/37] perf/x86/intel: Add support/quirk for the MISPREDICT bit on Knights Landing CPUs Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 26/37] netfilter: conntrack: timeout interface depend on CONFIG_NF_CONNTRACK_TIMEOUT Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 27/37] arch/hexagon: fix kernel/dma.c build warning Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 28/37] hexagon: modify ffs() and fls() to return int Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 29/37] arm64: jump_label.h: use asm_volatile_goto macro instead of "asm goto" Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 30/37] drm/amdgpu: fix error handling in amdgpu_cs_user_fence_chunk Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 31/37] r8169: Clear RTL_FLAG_TASK_*_PENDING when clearing RTL_FLAG_TASK_ENABLED Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 32/37] s390/qeth: use vzalloc for QUERY OAT buffer Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 33/37] s390/qeth: don't dump past end of unknown HW header Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 34/37] cifs: read overflow in is_valid_oplock_break() Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 35/37] xen/manage: don't complain about an empty value in control/sysrq node Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 36/37] xen: avoid crash in disable_hotplug_cpu Sasha Levin
2018-10-01 0:39 ` [PATCH AUTOSEL 4.14 37/37] xen: fix GCC warning and remove duplicate EVTCHN_ROW/EVTCHN_COL usage 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=20181001003850.147107-19-alexander.levin@microsoft.com \
--to=alexander.levin@microsoft.com \
--cc=bskeggs@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=stable@vger.kernel.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®