DMA_PMD大页IO缓冲池 · v1

[Re: [RFC: DMA_PMD 01/22] iommu/dma: introduce CONFIG_DMA_PMD and metadata table] On 10/3/26 2:22 PM, Luigi Rizzo wrote: > diff –git a/drivers/iommu/Kconfig b/drivers/iommu/Kconfig > index 6e07bd6946

添加DMA_PMD池生命周期（dma_pmd_pool_create/destroy）和页回收钩子__dma_pmd_free_page()，挂钩到__free_pages_prepare()；异步退役通过WQ_MEM_RECLAIM工作线程，RCU宽限期后释放到伙伴系统。

添加split_page_compound()，高阶版split_page()，将大阶页拆分为独立引用计数的复合页，结果冻结供池停放，分发时page_ref_unfreeze。用于DMA_PMD池切割PMD页。

添加DMA_PMD池块分配函数（dma_pmd_pool_alloc_bulk_node等），用split_page_compound()切割PMD块，冻结子页，分发时page_ref_unfreeze()，忽略不可选GFP标志，处理__GFP_ZERO。

为DMA_PMD池内存添加全局上限（dma_pmd_max_pages）和shrinker，限制空闲页数量；分配前检查上限，shrinker在回收时退役空闲页，池锁下分离、锁外退役。

在iommu_dma_cookie中嵌入dma_pmd_window，记录每域为DMA_PMD保留的IOVA范围，并暴露访问器（dma_pmd_dma_window等）和dma_info_to_prot()，为后续映射层准备。无功能变更。

添加dma_pmd_domain_release()，在iommu_put_dma_cookie()中调用，安全释放域映射：标记域draining、清除位图、等待srcu，防止竞争unmap访问已释放页表。

使用每域IOVA窗口映射DMA_PMD页，dma_pmd_dma_map_phys()在域位已设时O(1)返回IOVA，首次映射安装PDE_ORDER叶PTE。iommu_dma_map_phys/unmap_phys路由DMA_PMD缓冲区，使unmap为O(1)范围检查。

添加DMA_PMD arena分配器，用于dma_alloc_attrs()的长期设备区域。在物理RAM映射IOVA之后保留专用IOVA区域，便于通过IOVA识别arena分配。

Add four boolean attributes to struct device exposed under /sys/devices///dma_pmd_*: - dma_pmd_rxbuf: enable DMA_PMD pools for RX buffers / page_pool - dma_pmd_tx_hdrs: enable DMA_PMD pools fo

Wire dma_alloc_attrs() and dma_free_attrs() into the DMA_PMD arena when dev->dma_pmd_rings is enabled on a device: - Add dma_pmd_dma_alloc() and dma_is_pmd_dma() backed by dma_pmd_arena. - Rou

Add sysctl net.core.tx_enable_dma_pmd to back skb_page_frag_refill() with per-CPU order-3 (32KB block) and order-0 (4KB block) DMA_PMD page pools. When enabled, TX socket buffer page fragment allocat

When dev->dma_pmd_rxbuf is enabled on the pool device, use a per-page_pool DMA_PMD pool for page_pool allocations (__page_pool_alloc_netmems_slow() and __page_pool_alloc_page_order()). When enabled,

In Confidential Computing (CoCo) guests (AMD SEV-SNP, Intel TDX, ARM CCA), guest memory is private/encrypted by default and DMA buffers must be converted to shared/decrypted in the host page tables be

Add a background scrubber worker (dma_pmd_scrub_work) that zeroes dirty blocks in meta->dirty_bitmap[] and moves them to meta->free_bitmap[], and change __GFP_ZERO (want_init_on_alloc) allocations to

Add a per-NUMA-node reservoir of pre-allocated, pre-zeroed, and pre-prepared 2MB pages (dma_pmd_reservoirs[MAX_NUMNODES]) refilled in process context by a background worker (dma_pmd_reservoir_work):

When dev->dma_pmd_rxbuf is enabled, back GQI-RDA RX packet buffers with a per-queue DMA_PMD page pool (rx->dma_pmd_pool); DQO-RDA uses page_pool, which checks dev->dma_pmd_rxbuf directly. Signed-off-

When dev->dma_pmd_tx_hdrs is enabled, allocate a MAX_TCP_HEADER bounce buffer per TX slot/packet out of the DMA_PMD arena in raw-addressing mode (both GQI and DQO). When skb_headlen(skb) <= MAX_TCP_HE

When dev->dma_pmd_tx_hdrs is enabled, allocate a MAX_TCP_HEADER bounce buffer per TX DMA FIFO slot out of the DMA_PMD arena. When the linear portion of a transmitted packet is at most MAX_TCP_HEADER b

When dev->dma_pmd_tx_hdrs is enabled, allocate a MAX_TCP_HEADER bounce buffer per TX buffer slot out of the DMA_PMD arena. When the linear portion of a transmitted packet is at most MAX_TCP_HEADER byt

When dev->dma_pmd_tx_hdrs is enabled, allocate a MAX_TCP_HEADER bounce buffer per TX ring slot out of the DMA_PMD arena. When the linear portion of a transmitted packet is at most MAX_TCP_HEADER bytes

DMA_PMD pools fail silently. dma_pmd_pool_alloc() returns NULL for four distinct reasons - allocation backoff, the global DMA_PMD page ceiling, an PMD_ORDER allocation failure, and a split_page_compou