[PATCH] powerpc/pseries/iommu: switch to Default DMA window during kdump

In PowerPC (pseries) a non-virtualized adapter will have 2 DMA windows -
2GB default and a larger Dynamic DMA Window (DDW). DDW is large enough to
map total RAM to a device.

During normal functioning of OS, since RAM is pre-mapped, 2GB default
window is not used. The only scenario it might get used is when buffers in
pmemory are mapped to the device for DMA.

As of today, during kdump, during early device discovery, pci_dma_find()
finds that the device has 2 DMA windows. It selects to use DDW. This is a
kdump path and DMA window is needed for IO to the device.

Since, in the previous life of the LPAR (before panic), RAM was pre-mapped
via DDW, the DDW is completely full. So, in the iommu table initialization
code, iommu_table_clear() frees KDUMP_MIN_TCE_ENTRIES (2K) number of
TCEs.

But, it seems these are not enough for NVMe over Fibre-channel. When kdump
is trying to save vmcore on storage device, which is NVMe-FC, the
TCE usage is much more than 2K number of entries. The driver is mapping
a lot more buffers for DMA. After all the TCEs are consumed, iommu returns
iommu_alloc failures and the driver is not able to further map buffers for
IO. kdump fails to copy vmcore to NVMe-FC storage device.

Here are the driver logs and stack

lpfc 0153:70:00.0: iommu_alloc failed,
tbl 0000000034ebcf5e vaddr 00000000d814df0b npages 1
lpfc 0153:70:00.0: FCP Op failed - cmdiu dma mapping failed.
lpfc 0153:70:00.0: iommu_alloc failed,
tbl 0000000034ebcf5e vaddr 000000009779e4d2 npages 1
lpfc 0153:70:00.0: FCP Op failed - cmdiu dma mapping failed.

iommu_map_phys+0x1c4/0x1f0 (unreliable)
dma_iommu_map_phys+0x54/0xa0
dma_map_phys+0x3f8/0x590
__nvme_fc_init_request+0x110/0x300 [nvme_fc]
nvme_fc_init_request+0x60/0xb8 [nvme_fc]
blk_mq_alloc_map_and_rqs+0x388/0x510
blk_mq_alloc_tag_set+0x2a4/0x5f0
nvme_alloc_io_tag_set+0xe0/0x1e0 [nvme_core]
nvme_fc_connect_ctrl_work+0x85c/0xdac [nvme_fc]
process_one_work+0x1e4/0x5a0
worker_thread+0x1ec/0x3e0

Increasing the number of free TCE entries in iommu_table_clear() will
increase the probability of hitting EEH since there could still be some
active IOs from the previous life of the kernel.

Instead, during kdump, we can switch to default 2GB DMA window. This window
will mostly be empty. Or, could be slightly used if buffers in pmemory
were mapped for IO.

Fixes: 09a3c1e46142 ("powerpc/pseries/iommu: IOMMU table is not initialized for kdump over SR-IOV")
Signed-off-by: Gaurav Batra <redacted>
---
 arch/powerpc/platforms/pseries/iommu.c | 23 ++++++++++++-----------
 1 file changed, 12 insertions(+), 11 deletions(-)

diff --git a/arch/powerpc/platforms/pseries/iommu.c b/arch/powerpc/platforms/pseries/iommu.c
index 3e1f915fe4f6..272e9aab66d4 100644
--- a/arch/powerpc/platforms/pseries/iommu.c
+++ b/arch/powerpc/platforms/pseries/iommu.c

@@ -812,18 +812,11 @@ static struct device_node *pci_dma_find(struct device_node *dn,
 
 	/* parse DMA window property. During normal system boot, only default
 	 * DMA window is passed in OF. But, for kdump, a dedicated adapter might
-	 * have both default and DDW in FDT. In this scenario, DDW takes precedence
-	 * over default window.
+	 * have both default and DDW in FDT. In this scenario, default window
+	 * takes precedence over DDW. For a dedicated adapter, default window will
+	 * potentially have more unused TCEs.
 	 */
-	if (ddw_win) {
-		struct dynamic_dma_window_prop *p;
-
-		p = (struct dynamic_dma_window_prop *)ddw_prop;
-		prop->liobn = p->liobn;
-		prop->dma_base = p->dma_base;
-		prop->tce_shift = p->tce_shift;
-		prop->window_shift = p->window_shift;
-	} else if (default_win) {
+	if (default_win) {
 		unsigned long offset, size, liobn;
 
 		of_parse_dma_window(rdn, default_prop, &liobn, &offset, &size);

@@ -832,6 +825,14 @@ static struct device_node *pci_dma_find(struct device_node *dn,
 		prop->dma_base = cpu_to_be64(offset);
 		prop->tce_shift = cpu_to_be32(IOMMU_PAGE_SHIFT_4K);
 		prop->window_shift = cpu_to_be32(order_base_2(size));
+	} else {
+		struct dynamic_dma_window_prop *p;
+
+		p = (struct dynamic_dma_window_prop *)ddw_prop;
+		prop->liobn = p->liobn;
+		prop->dma_base = p->dma_base;
+		prop->tce_shift = p->tce_shift;
+		prop->window_shift = p->window_shift;
 	}
 
 	return rdn;

base-commit: 4549871118cf616eecdd2d939f78e3b9e1dddc48
-- 
2.39.3