MdeModulePkg/NonDiscoverablePciDeviceDxe: Avoid uncached bounce buffers #11722
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
The current handling of non-coherent DMA is very inefficient: - outbound DMA uses bounce buffers unnecessarily, considering that a bus master read cannot break coherency, and can therefore access the host buffer directly without the risk of data corruption; - bounce buffers are allocated from uncached memory, resulting in slow CopyMem() performance, and unnecessary overhead in allocating, remapping and freeing the pages, which also involves EFI memory map changes each time. So refactor the logic, and tweak it so that: - Bounce buffering is only performed when the host buffer is out of DMA range for the device, or when performing misaligned inbound DMA on a host buffer that is mapped cacheable; - Cache maintenance is only performed when the buffer presented to the device is mapped cacheable. - Bounce buffers are pool allocated, which is much more efficient and does not involve EFI memory map changes in the common case. Signed-off-by: Ard Biesheuvel <[email protected]>
Member
Author
os-d
reviewed
Nov 6, 2025
MdeModulePkg/Bus/Pci/NonDiscoverablePciDeviceDxe/NonDiscoverablePciDeviceIo.c
Show resolved
Hide resolved
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The current handling of non-coherent DMA is very inefficient:
outbound DMA uses bounce buffers unnecessarily, considering that a bus master read cannot break coherency, and can therefore access the host buffer directly without the risk of data corruption;
bounce buffers are allocated from uncached memory, resulting in slow CopyMem() performance, and unnecessary overhead in allocating, remapping and freeing the pages, which also involves EFI memory map changes each time.
So refactor the logic, and tweak it so that:
Bounce buffering is only performed when the host buffer is out of DMA range for the device, or when performing misaligned inbound DMA on a host buffer that is mapped cacheable;
Cache maintenance is only performed when the buffer presented to the device is mapped cacheable.
Bounce buffers are pool allocated, which is much more efficient and does not involve EFI memory map changes in the common case.