dma_iova_sync(9)
okay so bestie, welcome to dma_iova_sync(9) from the Kernel API Manual π«Ά this one is short but iconic, and we are gonna read it together, no cap β¨
π NAME
dma_iova_sync β Sync IOTLB
β¨ SYNOPSIS
int dma_iova_sync (struct device *dev , struct dma_iova_state *state ,
size_t offset , size_t size );
π ARGUMENTS
devβ DMA device. The device doing the DMA thing, basically the main character here πstateβ IOVA state. Yep, that state.offsetβ offset into the IOVA state to sync. It tells you where in the state we're syncing from.sizeβ size of the buffer. How big the buffer is, periodt.
π¦ DESCRIPTION
okay so the whole vibe: sync the IOTLB for the given IOVA state πβ¨
This function should be called on the IOVA-contiguous range created by one ore more dma_iova_link calls, and its job is to sync the IOTLB. Yes, βoreβ is literally what it says, and yes I am preserving it fr fr ππ
bestie, the key detail is that βshould be calledβ on that IOVA-contiguous range from one ore more dma_iova_link calls β that is the whole instruction, no cap.
πΈ SEE ALSO
Kernel file ./drivers/iommu/dma-iommu.c β go peek, iykyk π
iommu_get_dma_cookie(9)iommu_get_msi_cookie(9)iommu_put_dma_cookie(9)iommu_put_msi_cookie(9)iommu_dma_get_resv_regions(9)iommu_dma_init_options(9)iommu_dma_init_domain(9)dma_info_to_prot(9)dma_iova_try_alloc(9)dma_iova_free(9)dma_iova_link(9)dma_iova_unlink(9)dma_iova_destroy(9)
mwah π this lil pageβs source/version is ./drivers/iommu, and itβs dated October 2026 β iconic, stamped, done πβ¨
π¦ where this page comes from
- nixpkgs unstable: linux-manual 7.3.1
- π girlified by deepseek-ai/DeepSeek-V4.1-Flash-1017 on : a language model's rewrite of the original, under its terms