Home
Projects
Resources
Alternatives
Blog
Sign In
Hring Versions
Save
HTTP 1+2 in Rust, with io_uring & ktls
Overview
Versions
Reviews
Resources
httpwg-v0.1.1
2 weeks ago
Other
updated the following local packages: fluke-buffet, fluke-hpack
fluke-v0.1.1
2 weeks ago
Added
Upgrade dependencies
Fixed
fix more tests
fix several low-hanging bugs: re io/read tasks
Other
Try out pre-commit hook
More testing facilities
Make rfc9113 pass, still need to check codes
Alright, back to writing tests!
Introduce httpwg-macros, httpwg-gen
Woo, namespacing tests, getting rid of the useless struct. Maybe we can codegen the rest actually.
Woahey, 4.2 test works?
More test facilities
Flesh out the handshake test
4.1 test work
Well, well, almost got a working 4.1 test
Introduce 4.1 test (needs read loop in the background)
More test harness generation
Generate tests
Introduce start_server function
Use pipe for httpwg tests
Migrate ChanRead uses over to pipe
Finish up pipe implementation
Simplify WriteOwned trait, introduce pipe
First test is doing things
Introduce fluke-h2-parse
Some tests are starting to
almost
pass with io-uring-async
Fix buffer bug
Remove BufOrSlice, closes
#153
Bring 'fluke-maybe-uring' back into 'fluke-buffet'
Do a
bit
of flow control on incoming streams
Improve code coverage, closes
#151
Fix linux build
Implement final flow control touches
remove unused methods
6.9.1 all three cases pass
well
I
think 6.9.1 should pass.
Headers flow control wip
wip header flow control
do flow control accounting for headers as well
curl tests work! partial writes work.
wip Piece
Finish propagating Piece/PieceCore split changes
Piece::Slice
drat, partial writes.
Ready to write
a little forgetfulness
BodyEnd sets eof to true
Remove erroneous/outdated comment (the Frame struct no longer contains the payload)
Pass h2spec 6.9 tests apparently, even though we're never sending data now?
Introduce StreamIncoming, StreamOutgoing
Start implementing flow-control
hapsoc => fluke
Fix h2spec http2/6.4/3
Fix h2spec/6.5.2/1
Fix h2spec http2/6.5/3
More h2spec cases
Close connection explicitly in h2spec
Upgrade tokio-uring
mhh
Fix more cases
well every time I lose silly time to this I improve debug logging, so.
mh
Uhm
better debug implementation for Frame
Remove read module, move it back into server.rs
Remove write module altogether
Move all writing out of write
Retire H2ConnEvent
mhmh
Move settings acknowledgement inside read
Deprecate H2ConnEvent::RstStream
Move goaway writing away from write
Move ping to write_frame
Introduce write_frame
Give out_scratch to h2readcontext
Get rid of write task
Fix some flow-control cases
Fix PING tests
Fix more tests
Oh joy, h2spec is wrong
More test fixes
Fix more stream state cases
more stream state things
Fix some stream state cases
Rename StreamStage to StreamState
Implement RstStream
Fix hpack tests
looking good
No more compile errors woo
Simplifying headers/trailing reading + being more rigorous with conneciton errors
no continuation state
Upgrade more deps
Upgrade pretty-hex
Upgrade deps
Graceful GOAWAY handling
Set minimum rust version for crates with async fn in trait
Bump some dependencies
Remove TAIT
Bump dependencies
Opt out of async_fn_in_trait warnings
Move fluke-h2spec somewhere else
Move curl tests into their own crate
Simplify H2Result code further, try sccache.exe on Windows
Have a few methods return H2Result, which clarifies the control flow
Fix http2/5.1/5
Introduce 'write' module
Move types around
Bring state names closer to RFC
First stab at 5.1.2
todo => debug, don't crash on window_update for known streams
Don't send GO_AWAY when receiving a connection-wide window update
Send GO_AWAY if we get a stray WINDOW_UPDATE
Extract process_frame method out of H2ReadContext::work
release
fluke-hpack-v0.3.1
2 weeks ago
Added
Upgrade dependencies
Other
Fix buffer bug
hapsoc => fluke
Upgrade pretty-hex
Upgrade deps
Bump some dependencies
Bump dependencies
More drive-by clippy fixes
Make clippy happy by removing an .into_iter()
release
fluke-h2-parse-v0.1.1
2 weeks ago
Other
updated the following local packages: fluke-buffet
fluke-buffet-v0.2.0
2 weeks ago
Added
Upgrade dependencies
Other
Fix tests for Linux
Make rfc9113 pass, still need to check codes
progress
Flesh out the handshake test
Migrate more tests to pipe, deprecate ChanRead/ChanWrite
Migrate ChanRead uses over to pipe
Adjust shutdown implementation for TcpWriteHalf (io-uring codepath)
Port test to ChanRead
More assertions
Finish up pipe implementation
Simplify WriteOwned trait, introduce pipe
Cancel listen
Some cancellation?
Remove listen cancellation
Remove WrappedOp
wip cancellation
Only import io-uring on Linux, try closing fds on TcpStream drop
make io-uring dep linux-only
All tests pass!
Some tests are starting to
almost
pass with io-uring-async
flesh out TcpListener API
We can successfully accept a connection!
A little print debugging
Use nix for errno stuff
Vendor io-uring-async and upgrade it to io-uring 0.6.x
io-uring-async (cf.
#154
)
TcpWriteHalf fixes
Fix buffer bug
Remove BufOrSlice, closes
#153
Bring 'fluke-maybe-uring' back into 'fluke-buffet'
Fix linux build
Headers flow control wip
wip header flow control
wip Piece
Finish propagating Piece/PieceCore split changes
Piece::Slice
drat, partial writes.
hapsoc => fluke
Fix http2/4.2/1
Fix h2spec http2/6.5/3
mh
no continuation state
Upgrade more deps
Upgrade memmap2
Upgrade pretty-hex
Upgrade deps
Switch to Rust stable, closes
#128
Bump some dependencies
Bump dependencies
Opt out of async_fn_in_trait warnings
release
fluke-io-uring-async-v0.1.0
2 weeks ago
httpwg-macros-v0.1.0
2 weeks ago
httpwg-v0.1.0
2 weeks ago
fluke-h2-parse-v0.1.0
2 weeks ago
fluke-maybe-uring-v0.1.1
8 months ago
Other
Add more READMEs
« Previous
Next »
Home
Projects
Resources
Alternatives
Blog
Sign In
Sign In to OSA
I agree with
Terms of Service
and
Privacy Policy
Sign In with Github