Commit f3725813 authored by Moshe Shemesh's avatar Moshe Shemesh Committed by Jakub Kicinski
Browse files

net/mlx5: Ensure fw pages are always allocated on same NUMA 



When firmware asks the driver to allocate more pages, using event of
give_pages, the driver should always allocate it from same NUMA, the
original device NUMA. Current code uses dev_to_node() which can result
in different NUMA as it is changed by other driver flows, such as
mlx5_dma_zalloc_coherent_node(). Instead, use saved numa node for
allocating firmware pages.

Fixes: 311c7c71 ("net/mlx5e: Allocate DMA coherent memory on reader NUMA node")
Signed-off-by: default avatarMoshe Shemesh <moshe@nvidia.com>
Reviewed-by: default avatarTariq Toukan <tariqt@nvidia.com>
Signed-off-by: default avatarMark Bloch <mbloch@nvidia.com>
Link: https://patch.msgid.link/20250610151514.1094735-2-mbloch@nvidia.com


Signed-off-by: default avatarJakub Kicinski <kuba@kernel.org>
parent c7bdeed3

drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c

+1 −1
Original line number Diff line number Diff line
@@ -291,7 +291,7 @@ static void free_4k(struct mlx5_core_dev *dev, u64 addr, u32 function) 
static int alloc_system_page(struct mlx5_core_dev *dev, u32 function)

{

	struct device *device = mlx5_core_dma_dev(dev);

	int nid = dev_to_node(device);

	int nid = dev->priv.numa_node;

	struct page *page;

	u64 zero_addr = 1;

	u64 addr;