kernel.git/drivers/net/ethernet/google, branch linux-5.16.y

gve: Recording rx queue before sending to napi

2022-02-16T11:58:35Z

[ Upstream commit 084cbb2ec3af2d23be9de65fcc9493e21e265859 ] This caused a significant performance degredation when using generic XDP with multiple queues. Fixes: f5cedc84a30d2 ("gve: Add transmit and receive support") Signed-off-by: Tao Liu Link: https://lore.kernel.org/r/20220207175901.2486596-1-jeroendb@google.com Signed-off-by: Jakub Kicinski Signed-off-by: Sasha Levin

gve: fix the wrong AdminQ buffer queue index check

2022-02-08T17:35:17Z

commit 1f84a9450d75e08af70d9e2f2d5e1c0ac0c881d2 upstream. The 'tail' and 'head' are 'unsigned int' type free-running count, when 'head' is overflow, the 'int i (= tail) < u32 head' will be false: Only '- loop 0: idx = 63' result is shown, so it needs to use 'int' type to compare, it can handle the overflow correctly. typedef uint32_t u32; int main() { u32 tail, head; int stail, shead; int i, loop; tail = 0xffffffff; head = 0x00000000; for (i = tail, loop = 0; i < head; i++) { unsigned int idx = i & 63; printf("+ loop %d: idx = %u\n", loop++, idx); } stail = tail; shead = head; for (i = stail, loop = 0; i < shead; i++) { unsigned int idx = i & 63; printf("- loop %d: idx = %u\n", loop++, idx); } return 0; } Fixes: 5cdad90de62c ("gve: Batch AQ commands for creating and destroying queues.") Signed-off-by: Haiyue Wang Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman

gve: Fix GFP flags when allocing pages

2022-02-01T16:29:18Z

[ Upstream commit a92f7a6feeb3884c69c1c7c1f13bccecb2228ad0 ] Use GFP_ATOMIC when allocating pages out of the hotpath, continue to use GFP_KERNEL when allocating pages during setup. GFP_KERNEL will allow blocking which allows it to succeed more often in a low memory enviornment but in the hotpath we do not want to allow the allocation to block. Fixes: f5cedc84a30d2 ("gve: Add transmit and receive support") Signed-off-by: Catherine Sullivan Signed-off-by: David Awogbemila Link: https://lore.kernel.org/r/20220126003843.3584521-1-awogbemila@google.com Signed-off-by: Jakub Kicinski Signed-off-by: Sasha Levin

gve: Correct order of processing device options

2021-12-21T02:47:56Z

The legacy raw addressing device option was processed before the new RDA queue format option. This caused the supported features mask, which is provided only on the RDA queue format option, not to be set. This disabled jumbo-frame support when using raw adressing. Fixes: 255489f5b33c ("gve: Add a jumbo-frame device option") Signed-off-by: Jeroen de Borst Link: https://lore.kernel.org/r/20211220192746.2900594-1-jeroendb@google.com Signed-off-by: Jakub Kicinski

gve: fix for null pointer dereference.

2021-12-08T04:57:17Z

Avoid passing NULL skb to __skb_put() function call if napi_alloc_skb() returns NULL. Fixes: 37149e9374bf ("gve: Implement packet continuation for RX.") Signed-off-by: Ameer Hamza Link: https://lore.kernel.org/r/20211205183810.8299-1-amhamza.mgc@gmail.com Signed-off-by: Jakub Kicinski

gve: fix unmatched u64_stats_update_end()

2021-11-10T14:42:25Z

The u64_stats_update_end() call is supposed to be inside the curly braces so it pairs with the u64_stats_update_begin(). Fixes: 37149e9374bf ("gve: Implement packet continuation for RX.") Signed-off-by: Dan Carpenter Signed-off-by: David S. Miller

gve: Fix off by one in gve_tx_timeout()

2021-11-09T13:58:46Z

The priv->ntfy_blocks[] has "priv->num_ntfy_blks" elements so this > needs to be >= to prevent an off by one bug. The priv->ntfy_blocks[] array is allocated in gve_alloc_notify_blocks(). Fixes: 87a7f321bb6a ("gve: Recover from queue stall due to missed IRQ") Signed-off-by: Dan Carpenter Signed-off-by: David S. Miller

gve: Add a jumbo-frame device option.

2021-10-25T13:13:12Z

A widely deployed driver has a bug that will cause the driver not to load when a max_mtu > 2048 is present in the device descriptor. To avoid this bug while still enabling jumbo frames, we present a lower max_mtu in the device descriptor and pass the actual max_mtu in a separate device option. The driver supports 2 different queue formats. To enable features on one queue format, but not the other, a supported_features mask was added to the device options in the device descriptor. Signed-off-by: Shailend Chand Signed-off-by: Jeroen de Borst Signed-off-by: David S. Miller

gve: Implement packet continuation for RX.

2021-10-25T13:13:12Z

This enables the driver to receive RX packets spread across multiple buffers: For a given multi-fragment packet the "packet continuation" bit is set on all descriptors except the last one. These descriptors' payloads are combined into a single SKB before the SKB is handed to the networking stack. This change adds a "packet buffer size" notion for RX queues. The CreateRxQueue AdminQueue command sent to the device now includes the packet_buffer_size. We opt for a packet_buffer_size of PAGE_SIZE / 2 to give the driver the opportunity to flip pages where we can instead of copying. Signed-off-by: David Awogbemila Signed-off-by: Jeroen de Borst Reviewed-by: Catherine Sullivan Signed-off-by: David S. Miller

gve: Add RX context.

2021-10-25T13:13:12Z

This refactor moves the skb_head and skb_tail fields into a new gve_rx_ctx struct. This new struct will contain information about the current packet being processed. This is in preparation for multi-descriptor RX packets. Signed-off-by: David Awogbemila Signed-off-by: Jeroen de Borst Reviewed-by: Catherine Sullivan Signed-off-by: David S. Miller