Repository navigation
fix(psvr2): don't change an ioData buffer's length where macOS refuses - #7
Merged
Merged
Conversation
… to change it startReader set the length of each endpoint's ioDataWithCapacity buffer to its transfer size. Some macOS versions give such a buffer its capacity as its length and throw ioDataSetLengthException on any change, which ended metalvrd as soon as a PlayStation VR2 was found. The length is set only when it differs, the refusal is caught and logged, and a buffer shorter than the transfer leaves that endpoint unopened instead.
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
A PlayStation VR2 owner's
metalvrdended right after finding the headset:startReadersetr.data.length = sizeon the bufferioDataWithCapacity:returned. On their macOS that buffer already has its capacity as its length, and any change throws.Now the length is set only when it differs from the transfer size, inside
@try. A refusal is logged, and a buffer still shorter than the transfer leaves that endpoint unopened (startReaderreturns false, as for a pipe that doesn't open) instead of reading short. The Rift's USB shim (usb_shim.mm) never sets an ioData length, so it isn't affected.Tested
metalvrdbuilds (cmake --build build/native --target metalvrd).