There's a bug that when using the XEN hypervisor with dm-crypt on NVMe, the kernel deadlocks [1]. The deadlocks are caused by inability to map a large bio vector - dma_map_sgtable always returns an error, this gets propagated to the block layer as BLK_STS_RESOURCE and the block layer retries the request indefinitely. XEN uses the swiotlb framework to map discontiguous pages into contiguous runs that are submitted to the PCIe device. The swiotlb framework has a limitation on the length of a mapping - this needs to be announced with the max_mapping_size method to make sure that the hardware drivers do not create larger mappings. Without max_mapping_size, the NVMe block driver would create large mappings that overrun the maximum mapping size. [1] https://lore.kernel.org/stable/ZTNH0qtmint%2FzLJZ@mail-itl/ Signed-off-by: Mikulas Patocka <mpatocka@xxxxxxxxxx> Reported-by: Marek Marczykowski-Górecki <marmarek@xxxxxxxxxxxxxxxxxxxxxx> Tested-by: Marek Marczykowski-Górecki <marmarek@xxxxxxxxxxxxxxxxxxxxxx> Suggested-by: Keith Busch <kbusch@xxxxxxxxxx> Suggested-by: Christoph Hellwig <hch@xxxxxx> Cc: stable@xxxxxxxxxxxxxxx --- drivers/xen/swiotlb-xen.c | 1 + 1 file changed, 1 insertion(+) Index: linux-stable/drivers/xen/swiotlb-xen.c =================================================================== --- linux-stable.orig/drivers/xen/swiotlb-xen.c 2023-11-03 17:57:18.000000000 +0100 +++ linux-stable/drivers/xen/swiotlb-xen.c 2023-11-06 15:30:59.000000000 +0100 @@ -405,4 +405,5 @@ const struct dma_map_ops xen_swiotlb_dma .get_sgtable = dma_common_get_sgtable, .alloc_pages = dma_common_alloc_pages, .free_pages = dma_common_free_pages, + .max_mapping_size = swiotlb_max_mapping_size, };
