2220cf34a0
chore: rerun formatter
5 years ago
c4142c00f1
chore: more cleanup
5 years ago
0c85cd4102
chore: Assorted cleanup
5 years ago
862143f37a
chore: clean up interrupt logic
5 years ago
3d6266750e
chore: assorted refactors
5 years ago
63a6e702d9
chore: cleanup JSON parsing logic
5 years ago
69086d012d
Various refactors and end-to-end tracking of cycles
5 years ago
d058f99e0f
feat: Assign start_time to sandbox
5 years ago
b0434ac941
build: remove STANDALONE, SBOX_SCALE_ALLOC, USE_LIBUV flags
5 years ago
ae93435e2a
refactor: Assorted refactors and add start_time to request
5 years ago
6946b08644
refactor: modularize and rename things
5 years ago
565a03db5d
refactor: make resource limits function
5 years ago
b5e9abced5
chore: Assorted cleanup
5 years ago
9185852236
docs: Add documentation to code ( #7 )
5 years ago
e61be257f9
chore: add clang-format from Composite ( #5 )
...
* chore: add clang-format from Composite
* chore: cleanup JSON
* chore: add clang-format from Composite
* chore: stylistic corrections
* chore: ignore swap files and vscode conf
* chore: check semver of clang-format and require 9
5 years ago
bfbe848b4b
chore: cleanup JSON
5 years ago
a63af01ceb
Merge branch 'master' of https://github.com/phanikishoreg/awsm-Serverless-Framework
5 years ago
cc38020fde
code_benches, some require math lib
5 years ago
ed234f9ac5
default: async io for http req/resp. sync based io doesn't seem to work with huge data like 500k-1m I tested. They just timeout with concurrency
5 years ago
77ae44337c
buffer problem in work1m tests
5 years ago
0c92419978
fixed syscall based sync http-req/resp
5 years ago
b9d7278c54
fixes 99% latency problem here (but a problem with 1m data transfer test, it is broken.)
5 years ago
362e8b0a67
Merge branch 'master' of https://github.com/phanikishoreg/awsm-Serverless-Framework
5 years ago
da43b22407
work 1k to 1m functions
5 years ago
9528f65b32
Synchronous I/O for http request/response fixes 99%tile latencies
...
* Async I/O is being used for reading and writing http request and
response respectively. This causes the worker core to steal new tasks as
when one sandbox blocks on "read" or "write". This causes true
round-robin on timer interrupts, leading to those high latencies.
* Ideally, the request/response should be synchronous I belive because
we are expected to have "low"/near-realtime latencies and the read/writes
are only performed when a client connects, so ideally not block!
Plus, we expect only small data transfers (in KBs) so that should also
support synchronous request/response processing.
5 years ago
f8cc5edb89
fibonacci serverless fn in awsm
5 years ago
645d138787
for runtime with multi-sandboxing, globals are to be populated per sandbox, that was what causing random crashes!
5 years ago
69b1f94fe8
scalable sandbox creation in each runtime core
5 years ago
576bc4b086
sandbox alloc on each core, for scalability
5 years ago
3fab99424a
fibonacci test update
5 years ago
affe5fbc35
read() chunks fix, other standalone compilation fixes
5 years ago
837fc3a7c0
legacy function pointer typecasts fail in wasm
5 years ago
5016942e3f
work test
5 years ago
611ec2257a
keep empty and add work program
5 years ago
ae392fdfe3
read from stdin (http req body), write to stdout (http resp body), bug fixes in http parsing
5 years ago
f019193952
read() to read from request body!
5 years ago
2e8a22b710
fix for standalone execution
5 years ago
be01f87687
Merge branch 'master' of https://github.com/phanikishoreg/awsm-Serverless-Framework
5 years ago
5968495a55
adds floor for float and double
5 years ago
618f269a9a
content-length header in response
5 years ago
27b317116d
reset offset in read
5 years ago
aa0017f4b0
http request response using uv
5 years ago
8a3e050282
Support TCP (TODO: http parsing based on module specification)
...
* Major:
- Moved sandbox data struct into linear memory of a sandbox.
- (Should work well with other sandboxing mechanisms).
- Removed all mallocs from within uv usage, using a pre-allocated
memory within sandbox data-struct for read/write.
- TODO: convert prints on stdout to a string and send it out with
function response body.
- TODO: for each function loaded, what is the request and response
http header/body specification.
To start, keep headers minimal.
body either: json, plain.
- TODO: track TIMEOUTS. If we go beyond that, return a HTTP 408?
5 years ago
9654c1f2fc
libc reuse mem for recv
5 years ago
d31057ebde
cleanup commented code
5 years ago
69b6dcf83c
added http-parser
5 years ago
9376453058
uvio for sandbox response.
5 years ago
95bf6fc1a9
udp response (return value of main) and multi-threaded udp client
5 years ago
eb8fcf00c4
Work-stealing deque for global request queue
...
* all cores steal from that queue.
* other bug fixes
* tested with NCORES = 2 and 4.
5 years ago
2f3ddcb9c5
fix tests makefile
5 years ago
670e3ad2b5
revert from standalone
5 years ago
a231796990
for standalone execution
5 years ago
7ce5cf9a6d
fix in wasm_timespec
5 years ago
9b0ec6f7dc
Changed directory structure
...
* silverfish is a submodule now
* memory modules from silverfish runtime is copied in to the aWsm runtime.
* runtime/tests/Makefile compiles different tests and silverfish/code_benches..
5 years ago
Sean McBride
phani