From: Oded Gabbay <ogabbay@kernel.org>
To: linux-kernel@vger.kernel.org
Subject: [PATCH 28/30] habanalabs: fix race between wait and irq
Date: Sat, 22 Jan 2022 21:57:29 +0200 [thread overview]
Message-ID: <20220122195731.934494-28-ogabbay@kernel.org> (raw)
In-Reply-To: <20220122195731.934494-1-ogabbay@kernel.org>
There is a race in the user interrupts code, where between checking
the target value and adding the new pend to the list, there is a chance
the interrupt happened.
In that case, no one will complete the node, and we will get a timeout
on it.
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
---
drivers/misc/habanalabs/common/command_submission.c | 6 +++++-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/drivers/misc/habanalabs/common/command_submission.c b/drivers/misc/habanalabs/common/command_submission.c
index 2b27ff826e3c..3d9020825df5 100644
--- a/drivers/misc/habanalabs/common/command_submission.c
+++ b/drivers/misc/habanalabs/common/command_submission.c
@@ -2901,16 +2901,21 @@ static int _hl_interrupt_wait_ioctl(struct hl_device *hdev, struct hl_ctx *ctx,
pend->cq_kernel_addr = (u64 *) cb->kernel_address + cq_counters_offset;
pend->cq_target_value = target_value;
+ spin_lock_irqsave(&interrupt->wait_list_lock, flags);
+
/* We check for completion value as interrupt could have been received
* before we added the node to the wait list
*/
if (*pend->cq_kernel_addr >= target_value) {
+ spin_unlock_irqrestore(&interrupt->wait_list_lock, flags);
+
*status = HL_WAIT_CS_STATUS_COMPLETED;
/* There was no interrupt, we assume the completion is now. */
pend->fence.timestamp = ktime_get();
goto set_timestamp;
} else if (!timeout_us) {
+ spin_unlock_irqrestore(&interrupt->wait_list_lock, flags);
*status = HL_WAIT_CS_STATUS_BUSY;
pend->fence.timestamp = ktime_get();
goto set_timestamp;
@@ -2919,7 +2924,6 @@ static int _hl_interrupt_wait_ioctl(struct hl_device *hdev, struct hl_ctx *ctx,
/* Add pending user interrupt to relevant list for the interrupt
* handler to monitor
*/
- spin_lock_irqsave(&interrupt->wait_list_lock, flags);
list_add_tail(&pend->wait_list_node, &interrupt->wait_list_head);
spin_unlock_irqrestore(&interrupt->wait_list_lock, flags);
--
2.25.1
next prev parent reply other threads:[~2022-01-22 19:59 UTC|newest]
Thread overview: 31+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-01-22 19:57 [PATCH 01/30] habanalabs: check the return value of hl_cs_poll_fences() Oded Gabbay
2022-01-22 19:57 ` [PATCH 02/30] habanalabs: fix race when waiting on encaps signal Oded Gabbay
2022-01-22 19:57 ` [PATCH 03/30] habanalabs: fix possible memory leak in MMU DR fini Oded Gabbay
2022-01-22 19:57 ` [PATCH 04/30] habanalabs/gaudi: disable CGM permanently Oded Gabbay
2022-01-22 19:57 ` [PATCH 05/30] habanalabs: remove ASIC functions of clock gating Oded Gabbay
2022-01-22 19:57 ` [PATCH 06/30] habanalabs: make some MMU functions common Oded Gabbay
2022-01-22 19:57 ` [PATCH 07/30] habanalabs: sysfs functions should be in sysfs.c Oded Gabbay
2022-01-22 19:57 ` [PATCH 08/30] habanalabs: get clk is common function Oded Gabbay
2022-01-22 19:57 ` [PATCH 09/30] habanalabs: remove hwmgr.c Oded Gabbay
2022-01-22 19:57 ` [PATCH 10/30] habanalabs: move more f/w functions to firmware_if.c Oded Gabbay
2022-01-22 19:57 ` [PATCH 11/30] habanalabs: remove asic callback set_pll_profile() Oded Gabbay
2022-01-22 19:57 ` [PATCH 12/30] habanalabs: rename dev_attr_grp to dev_clk_attr_grp Oded Gabbay
2022-01-22 19:57 ` [PATCH 13/30] habanalabs: add vrm version to sysfs Oded Gabbay
2022-01-22 19:57 ` [PATCH 14/30] habanalabs: remove power9 workaround for dma support Oded Gabbay
2022-01-22 19:57 ` [PATCH 15/30] habanalabs: use common wrapper for MMU cache invalidation Oded Gabbay
2022-01-22 19:57 ` [PATCH 16/30] habanalabs: sysfs support for fw os version Oded Gabbay
2022-01-22 19:57 ` [PATCH 17/30] habanalabs: there is no kernel TDR in future ASICs Oded Gabbay
2022-01-22 19:57 ` [PATCH 18/30] habanalabs: duplicate HOP table props to MMU props Oded Gabbay
2022-01-22 19:57 ` [PATCH 19/30] habanalabs: don't free phys_pg_pack inside lock Oded Gabbay
2022-01-22 19:57 ` [PATCH 20/30] habanalabs: avoid copying pll data if pll_info_get fails Oded Gabbay
2022-01-22 19:57 ` [PATCH 21/30] habanalabs: add missing error check in sysfs clk_freq_mhz_show Oded Gabbay
2022-01-22 19:57 ` [PATCH 22/30] habanalabs: fix soft reset flow in case of failure Oded Gabbay
2022-01-22 19:57 ` [PATCH 23/30] habanalabs: add missing error check in sysfs max_power_show Oded Gabbay
2022-01-22 19:57 ` [PATCH 24/30] habanalabs: update to latest f/w specs Oded Gabbay
2022-01-22 19:57 ` [PATCH 25/30] habanalabs: expose number of user interrupts Oded Gabbay
2022-01-22 19:57 ` [PATCH 26/30] habanalabs: reject host map with mmu disabled Oded Gabbay
2022-01-22 19:57 ` [PATCH 27/30] habanalabs: fix user interrupt wait when timeout is 0 Oded Gabbay
2022-01-22 19:57 ` Oded Gabbay [this message]
2022-01-22 19:57 ` [PATCH 29/30] habanalabs: prevent false heartbeat failure during soft-reset Oded Gabbay
2022-01-22 19:57 ` [PATCH 30/30] habanalabs: remove duplicate print Oded Gabbay
[not found] ` <20220123002722.3057-1-hdanton@sina.com>
2022-01-24 18:22 ` [PATCH 02/30] habanalabs: fix race when waiting on encaps signal Dani Liberman
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=20220122195731.934494-28-ogabbay@kernel.org \
--to=ogabbay@kernel.org \
--cc=linux-kernel@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;
as well as URLs for NNTP newsgroup(s).