summaryrefslogtreecommitdiffhomepage
path: root/pkg/buffer
AgeCommit message (Collapse)Author
2021-05-14Merge release-20210510.0-35-g2ac6b7688 (automated)gVisor bot
2021-05-14pkg/buffer: Remove dependency to safemem, code no longer usedTing-Yu Wang
PiperOrigin-RevId: 373846881
2021-05-13Merge release-20210510.0-27-g84f04cc85 (automated)gVisor bot
2021-05-13Migrate PacketBuffer to use pkg/bufferTing-Yu Wang
Benchmark iperf3: Before After native->runsc 5.14 5.01 (Gbps) runsc->native 4.15 4.07 (Gbps) It did introduce overhead, mainly at the bridge between pkg/buffer and VectorisedView, the ExtractVV method. Once endpoints start migrating away from VV, this overhead will be gone. Updates #2404 PiperOrigin-RevId: 373651666
2021-01-15Merge release-20210112.0-25-ge57ebcd37 (automated)gVisor bot
2021-01-11Merge release-20201216.0-82-g4c4de6644 (automated)gVisor bot
2020-10-09Merge release-20200928.0-78-g743327817 (automated)gVisor bot
2020-10-08Merge release-20200928.0-66-ga55bd73d4 (automated)gVisor bot
2020-09-23Merge release-20200914.0-135-gc0f21bb19 (automated)gVisor bot
2020-09-22pkg/buffer: Reorganize internal structure to allow dynamic sizes.Ting-Yu Wang
This change changes `buffer.data` into a `[]byte`, from `[bufferSize]byte`. In exchange, each `buffer` is now grouped together to reduce the number of allocation. Plus, `View` now holds an embeded list of `buffer` (via `pool`) to support the happy path which the number of buffer is small. Expect no extra allocation for the happy path. It is to enable the use case for PacketBuffer, which * each `View` is small (way less than `defaultBufferSize`), and * needs to dynamically transfer ownership of `[]byte` to `View`. (to allow gradual migration) PiperOrigin-RevId: 333197252
2020-08-27Fix vfs2 pipe behavior when splicing to a non-pipe.Dean Deng
Fixes *.sh Java runtime tests, where splice()-ing from a pipe to /dev/zero would not actually empty the pipe. There was no guarantee that the data would actually be consumed on a splice operation unless the output file's implementation of Write/PWrite actually called VFSPipeFD.CopyIn. Now, whatever bytes are "written" are consumed regardless of whether CopyIn is called or not. Furthermore, the number of bytes in the IOSequence for reads is now capped at the amount of data actually available. Before, splicing to /dev/zero would always return the requested splice size without taking the actual available data into account. This change also refactors the case where an input file is spliced into an output pipe so that it follows a similar pattern, which is arguably cleaner anyway. Updates #3576. PiperOrigin-RevId: 328843954
2020-06-24Merge release-20200608.0-119-g364ac92ba (automated)gVisor bot
2020-05-27Merge release-20200518.0-45-g0bc022b7 (automated)gVisor bot
2020-05-26Implement splice(2) and tee(2) for VFS2.Jamie Liu
Updates #138 PiperOrigin-RevId: 313326354
2020-05-07Merge release-20200422.0-51-g1f4087e (automated)gVisor bot
2020-04-04Merge release-20200323.0-69-gfc99a7e (automated)gVisor bot
2020-03-12Merge release-20200219.0-153-g61051f2 (automated)gVisor bot
2020-03-11Clean-up buffer implementation.Adin Scannell
This also adds substantial test cases. The Read/Write interfaces are dropped as they are not necessary. PiperOrigin-RevId: 300461547
2020-03-06Merge release-20200219.0-122-g6fa5cee (automated)gVisor bot
2020-03-03Merge release-20200219.0-91-g844e4d2 (automated)gVisor bot
2020-02-28Merge release-20200219.0-69-g463f421 (automated)gVisor bot
2020-02-28Make pipe buffer implementation standard.Adin Scannell
A follow-up change will convert the networking code to use this standard pipe implementation. PiperOrigin-RevId: 297903206