u-boot.git/drivers/nvme, branch v2026.07

drivers: nvme: Log I/O timeouts

2026-06-03T16:22:18Z

Current code silently swallows any timed-out commands scheduled to NVMe. Log those to be able to debug any potential problems with the NVMe hardware/firmware. Signed-off-by: Denis Mukhin Reviewed-by: Neil Armstrong Reviewed-by: Simon Glass Link: https://patch.msgid.link/20260529034441.2075305-2-dmukhin@ford.com Signed-off-by: Neil Armstrong

nvme: free prp_pool on nvme_init() failure paths

2026-05-28T10:56:52Z

nvme_init() allocates prp_pool after configuring the admin queue, but some later error paths return without freeing it. Free prp_pool before freeing the queue array in the failure paths after nvme_setup_io_queues() and namespace ID buffer allocation. This fixes a memory leak during NVMe initialization failures. Signed-off-by: Prashant Kamble Reviewed-by: Neil Armstrong Link: https://patch.msgid.link/20260524145721.9206-1-prashant.kamble223@gmail.com Signed-off-by: Neil Armstrong

nvme: avoid deleting uncreated queues

2026-05-28T10:56:23Z

nvme_create_queue() may issue Delete CQ or Delete SQ commands even when the corresponding queue creation failed. Avoid sending delete commands for queues that were never successfully created. Signed-off-by: Prashant Kamble Reviewed-by: Neil Armstrong Link: https://patch.msgid.link/20260524154718.16381-1-prashant.kamble223@gmail.com Signed-off-by: Neil Armstrong

nvme: fix dcache invalidation range in identify command

2026-05-28T10:56:02Z

When the identify buffer crosses a page boundary, PRP2 is used and dma_addr is advanced to the second page: dma_addr += (page_size - offset); The subsequent invalidate_dcache_range() calls then use the modified dma_addr instead of the original buffer start address. As a result, the beginning of the identify buffer is not invalidated and the invalidation range extends past the end of the buffer. Fix this by preserving the original DMA buffer address for cache invalidation. Suggested-by: Neil Armstrong Signed-off-by: Prashant Kamble Reviewed-by: Neil Armstrong Link: https://patch.msgid.link/20260524100625.11135-1-prashant.kamble223@gmail.com Signed-off-by: Neil Armstrong

nvme: Fix PRP list pointer arithmetic for chained transfers

2026-05-20T07:51:44Z

The PRP setup code advances prp_pool using u64 pointer arithmetic: prp_pool += page_size; This increments the pointer by page_size * sizeof(u64) bytes instead of page_size bytes, resulting in invalid PRP list addresses when multiple PRP list pages are required. The issue becomes visible for large transfers, typically above 2 MiB when MDTS > 9. Fix it by using byte-wise pointer arithmetic when advancing to the next PRP list page. Signed-off-by: Prashant Kamble Reviewed-by: Neil Armstrong Link: https://patch.msgid.link/20260518022535.17197-1-prashant.kamble223@gmail.com Signed-off-by: Neil Armstrong

nvme: Staticize and constify driver ops

2026-05-20T07:44:37Z

Set the ops structure as static const. The structure is not accessible from outside of this driver and is not going to be modified at runtime. Signed-off-by: Marek Vasut Reviewed-by: Neil Armstrong Link: https://patch.msgid.link/20260508122128.512798-1-marek.vasut+renesas@mailbox.org Signed-off-by: Neil Armstrong

nvme: apple: Check memalign return value

2026-05-20T07:44:37Z

memalign returns NULL if it fails. This commit ensures that we handle this failure before filling the buffer with 0s. Signed-off-by: Francois Berder Reviewed-by: Neil Armstrong Link: https://patch.msgid.link/BESP194MB280542535B098A33C8A815EEDA3A2@BESP194MB2805.EURP194.PROD.OUTLOOK.COM Signed-off-by: Neil Armstrong

nvme: fix command ID wraparound handling

2026-05-20T07:44:37Z

nvme_get_cmd_id() returns 0 after cmdid reaches USHRT_MAX, but fails to reset cmdid itself. As a result, all subsequent calls keep returning 0 indefinitely. Reset cmdid when wraparound occurs so command IDs continue incrementing correctly. Signed-off-by: Prashant Kamble Reviewed-by: Neil Armstrong Link: https://patch.msgid.link/20260518060915.45607-1-prashant.kamble223@gmail.com Signed-off-by: Neil Armstrong

arm: apple: Switch to board based text env

2026-05-12T18:08:15Z

The main use case for u-boot on Apple silicon based devices is to provide an EFI based bootloader for operating systems. This uses a generic u-boot image with DTBs passed from an earlier boot loader (m1n1). Use the generic board name "mac" for this purpose. Signed-off-by: Janne Grunau

nvme: apple: add "apple,t8103-nvme-ans2" compatible

2026-05-12T18:07:49Z

After discussion with the devicetree maintainers we agreed to not extend lists with the generic compatible "apple,nvme-ans2" anymore [1]. Add "apple,t8103-nvme-ans2" as fallback compatible as this is the SoC the driver and bindings were originally written for. The t602x (M2 Pro/Max/Ultra) devicetrees submitted in [2] use this compatible as fallback instead of "apple,t8103-nvme-ans2". Link: https://lore.kernel.org/asahi/12ab93b7-1fc2-4ce0-926e-c8141cfe81bf@kernel.org/ [1] Link: https://lore.kernel.org/asahi/20250828-dt-apple-t6020-v1-0-507ba4c4b98e@jannau.net/ [2] Signed-off-by: Janne Grunau Acked-by: Mark Kettenis