Received-SPF: pass (google.com: domain of linux-kernel+bounces-11790-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.80.249 as permitted sender) client-ip=147.75.80.249;
From: Ethan Zhao <haifeng.zhao@linux.intel.com>
To: bhelgaas@google.com,
	baolu.lu@linux.intel.com,
	dwmw2@infradead.org,
	will@kernel.org,
	robin.murphy@arm.com,
	lukas@wunner.de
Cc: linux-pci@vger.kernel.org,
	iommu@lists.linux.dev,
	linux-kernel@vger.kernel.org
Subject: [RFC PATCH v8 5/5] iommu/vt-d: don't loop for timeout device-TLB invalidation request forever
Date: Tue, 26 Dec 2023 21:59:23 -0500
Message-Id: <20231227025923.536148-6-haifeng.zhao@linux.intel.com>
In-Reply-To: <20231227025923.536148-1-haifeng.zhao@linux.intel.com>
References: <20231227025923.536148-1-haifeng.zhao@linux.intel.com>
Precedence: bulk
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit

When the device-TLB invalidation (ATS invalidation) timeout happens, the
qi_submit_sync() will restart and loop for the invalidation request
forever till it is done, it will block another invalidation thread such
as the fq_timer to issue invalidation request, cause the system lockup as
following

[exception RIP: native_queued_spin_lock_slowpath+92]

RIP: ffffffffa9d1025c RSP: ffffb202f268cdc8 RFLAGS: 00000002

RAX: 0000000000000101 RBX: ffffffffab36c2a0 RCX: 0000000000000000

RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffffffffab36c2a0

RBP: ffffffffab36c2a0 R8: 0000000000000001 R9: 0000000000000000

R10: 0000000000000010 R11: 0000000000000018 R12: 0000000000000000

R13: 0000000000000004 R14: ffff9e10d71b1c88 R15: ffff9e10d71b1980

ORIG_RAX: ffffffffffffffff CS: 0010 SS: 0018

--- ---

(the left part of exception see the hotplug case of ATS capable device)

If one endpoint device just no response to the device-TLB invalidation
request, but is not gone, it will bring down the whole system, to avoid
such case, don't try the timeout device-TLB request forever.

and as synchronous program model of current qi_submit_sync() implementation
we couldn't wait for the enough time as PCIe spec said 1min+50%, just break
it in current sync model. (PCIe spec r6.1, sec 10.3.1)

Signed-off-by: Ethan Zhao <haifeng.zhao@linux.intel.com>
---
 drivers/iommu/intel/dmar.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/iommu/intel/dmar.c b/drivers/iommu/intel/dmar.c
index 76903a8bf963..206ab0b7294f 100644
--- a/drivers/iommu/intel/dmar.c
+++ b/drivers/iommu/intel/dmar.c
@@ -1457,7 +1457,7 @@ int qi_submit_sync(struct intel_iommu *iommu, struct qi_desc *desc,
 	reclaim_free_desc(qi);
 	raw_spin_unlock_irqrestore(&qi->q_lock, flags);
 
-	if (rc == -EAGAIN)
+	if (rc == -EAGAIN && type !=QI_DIOTLB_TYPE && type != QI_DEIOTLB_TYPE)
 		goto restart;
 
 	if (iotlb_start_ktime)
-- 
2.31.1