linux-kernel - [PATCH 1/1] iommu: dmar_fault should only clear PPF/PFO field.

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Message-Id: <1364199652-10090-1-git-send-email-zhen-hual@hp.com>
Date:	Mon, 25 Mar 2013 16:20:52 +0800
From:	"Li, Zhen-Hua" <zhen-hual@...com>
To:	lizhenhua@...com, Joerg Roedel <joro@...tes.org>,
	Ingo Molnar <mingo@...nel.org>,
	Donald Dutile <ddutile@...hat.com>,
	Hannes Reinecke <hare@...e.de>, linux-kernel@...r.kernel.org
Cc:	"Li, Zhen-Hua" <zhen-hual@...com>
Subject: [PATCH 1/1] iommu: dmar_fault should only clear PPF/PFO field.

When there is a dmar irq, dmar_fault is called and all of the fields 
in FSTS are cleared. But ICE/IQE/ITE should not be cleared here, 
they need to be processed and cleared in function qi_check_fault.

Signed-off-by: Li, Zhen-Hua <zhen-hual@...com>
---
 drivers/iommu/dmar.c |   14 +++++++++-----
 1 file changed, 9 insertions(+), 5 deletions(-)

diff --git a/drivers/iommu/dmar.c b/drivers/iommu/dmar.c
index e5cdaf8..d620117 100644
--- a/drivers/iommu/dmar.c
+++ b/drivers/iommu/dmar.c
@@ -1204,7 +1204,7 @@ irqreturn_t dmar_fault(int irq, void *dev_id)
 
 	/* TBD: ignore advanced fault log currently */
 	if (!(fault_status & DMA_FSTS_PPF))
-		goto clear_rest;
+		goto unlock_exit;
 
 	fault_index = dma_fsts_fault_record_index(fault_status);
 	reg = cap_fault_reg_offset(iommu->cap);
@@ -1245,11 +1245,12 @@ irqreturn_t dmar_fault(int irq, void *dev_id)
 			fault_index = 0;
 		raw_spin_lock_irqsave(&iommu->register_lock, flag);
 	}
-clear_rest:
-	/* clear all the other faults */
-	fault_status = readl(iommu->reg + DMAR_FSTS_REG);
-	writel(fault_status, iommu->reg + DMAR_FSTS_REG);
 
+	writel(DMA_FSTS_PPF, iommu->reg + DMAR_FSTS_REG);
+	if (fault_status & DMA_FSTS_PFO)
+		writel(DMA_FSTS_PFO, iommu->reg + DMAR_FSTS_REG);
+
+unlock_exit:
 	raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
 	return IRQ_HANDLED;
 }
@@ -1297,6 +1298,7 @@ int __init enable_drhd_fault_handling(void)
 	for_each_drhd_unit(drhd) {
 		int ret;
 		struct intel_iommu *iommu = drhd->iommu;
+		u32 fault_status;
 		ret = dmar_set_interrupt(iommu);
 
 		if (ret) {
@@ -1309,6 +1311,8 @@ int __init enable_drhd_fault_handling(void)
 		 * Clear any previous faults.
 		 */
 		dmar_fault(iommu->irq, iommu);
+		fault_status = readl(iommu->reg + DMAR_FSTS_REG);
+		writel(fault_status, iommu->reg + DMAR_FSTS_REG);
 	}
 
 	return 0;
-- 
1.7.10.4

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/