Commits
- Commit:
3a2c53f506aa2d07eaf0e8540f87054c622304fb
- From:
- Omar Polo <op@omarpolo.com>
- Date:
implement Tstat
- Commit:
d9985b95c77fcdb2d5878cd0b9e213a66123029c
- From:
- Omar Polo <op@omarpolo.com>
- Date:
drop extra do_send
- Commit:
6ace3b52f967ae34f9d1cf506669d4a2617bf115
- From:
- Omar Polo <op@omarpolo.com>
- Date:
fix typo. spotted by cage, thanks!
- Commit:
2483be55a471524cc46605fb83bd3a8a227d7d07
- From:
- Omar Polo <op@omarpolo.com>
- Date:
implement Twrite
- Commit:
fcf8ca156c415ee0c3c1b224205e09bddfad562e
- From:
- Omar Polo <op@omarpolo.com>
- Date:
allow directory to be read again using offset 0
The documentation says explicitly that the offset is *either* 0 or the
last offset plus count; that means that after some read a client can
require *again* an offset 0 (i.e. doing a rewinding the directory.)
- Commit:
5585f1c3803e04ba1341fef0f031b1df1312682a
- From:
- Omar Polo <op@omarpolo.com>
- Date:
implement Tread against directories
To do so it was needed to modify the np_* functions so they take an
explicit evbuffer.
- Commit:
20c188cec87a667632239f58cf43696ddb60ec77
- From:
- Omar Polo <op@omarpolo.com>
- Date:
PATH_MAX already accounts for NUL
- Commit:
1e8d72fd3f333ff75989cca10a0ae9285f4b5f2f
- From:
- Omar Polo <op@omarpolo.com>
- Date:
free evp only if it was allocated
apparently, evbuffer_free doesn't like NULLs
- Commit:
5f362b7f054ee9e876d7b8c208fcc09b2235091f
- From:
- Omar Polo <op@omarpolo.com>
- Date:
typo spotted by cage, thanks!
- Commit:
c0ae57b900c0dda41d12793b7a0480159f2946ca
- From:
- Omar Polo <op@omarpolo.com>
- Date:
initialize evbuffer only for dirs
- Commit:
baa405044968a73fd6a257a3504efe7dc5fe7443
- From:
- Omar Polo <op@omarpolo.com>
- Date:
first half of the tread implementation done, no new tests yet
- Commit:
d4e438158258d53a0d22189b432537b22300dd2e
- From:
- Omar Polo <op@omarpolo.com>
- Date:
allow opening directories (for reading only)
- Commit:
99a435901454cc6ca529a783486c83ffddadb110
- From:
- Omar Polo <op@omarpolo.com>
- Date:
fix walk fid check: look at fd not iomode to see if it was opened
- Commit:
87a818e80dd7b59d2ee3f9b1cdffc82d7c5c1b4c
- From:
- Omar Polo <op@omarpolo.com>
- Date:
ensure we don't call openat(2) with "" as path
- Commit:
cf758c3329364c1516e2b78c573d6901843bc670
- From:
- Omar Polo <op@omarpolo.com>
- Date:
fix Tattach handling
The previous implementation assumed that you can't attach more than
once. This is clearly wrong, Tattach and Twalk are the two ways to
obtain new fids.
This drops the error on subsequential attach and making the test
"multiple attach" passes.