[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20200213151859.926436608@linuxfoundation.org>
Date: Thu, 13 Feb 2020 07:21:00 -0800
From: Greg Kroah-Hartman <gregkh@...uxfoundation.org>
To: linux-kernel@...r.kernel.org
Cc: Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
stable@...r.kernel.org, Ram Pai <linuxram@...ibm.com>,
Alexey Kardashevskiy <aik@...abs.ru>,
Thiago Jung Bauermann <bauerman@...ux.ibm.com>,
Michael Ellerman <mpe@...erman.id.au>
Subject: [PATCH 5.4 53/96] Revert "powerpc/pseries/iommu: Dont use dma_iommu_ops on secure guests"
From: Ram Pai <linuxram@...ibm.com>
commit d862b44133b7a1d7de25288e09eabf4df415e971 upstream.
This reverts commit edea902c1c1efb855f77e041f9daf1abe7a9768a.
At the time the change allowed direct DMA ops for secure VMs; however
since then we switched on using SWIOTLB backed with IOMMU (direct mapping)
and to make this work, we need dma_iommu_ops which handles all cases
including TCE mapping I/O pages in the presence of an IOMMU.
Fixes: edea902c1c1e ("powerpc/pseries/iommu: Don't use dma_iommu_ops on secure guests")
Signed-off-by: Ram Pai <linuxram@...ibm.com>
[aik: added "revert" and "fixes:"]
Signed-off-by: Alexey Kardashevskiy <aik@...abs.ru>
Reviewed-by: Thiago Jung Bauermann <bauerman@...ux.ibm.com>
Tested-by: Thiago Jung Bauermann <bauerman@...ux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@...erman.id.au>
Link: https://lore.kernel.org/r/20191216041924.42318-2-aik@ozlabs.ru
Signed-off-by: Greg Kroah-Hartman <gregkh@...uxfoundation.org>
---
arch/powerpc/platforms/pseries/iommu.c | 11 +----------
1 file changed, 1 insertion(+), 10 deletions(-)
--- a/arch/powerpc/platforms/pseries/iommu.c
+++ b/arch/powerpc/platforms/pseries/iommu.c
@@ -36,7 +36,6 @@
#include <asm/udbg.h>
#include <asm/mmzone.h>
#include <asm/plpar_wrappers.h>
-#include <asm/svm.h>
#include "pseries.h"
@@ -1320,15 +1319,7 @@ void iommu_init_early_pSeries(void)
of_reconfig_notifier_register(&iommu_reconfig_nb);
register_memory_notifier(&iommu_mem_nb);
- /*
- * Secure guest memory is inacessible to devices so regular DMA isn't
- * possible.
- *
- * In that case keep devices' dma_map_ops as NULL so that the generic
- * DMA code path will use SWIOTLB to bounce buffers for DMA.
- */
- if (!is_secure_guest())
- set_pci_dma_ops(&dma_iommu_ops);
+ set_pci_dma_ops(&dma_iommu_ops);
}
static int __init disable_multitce(char *str)
Powered by blists - more mailing lists