Joey Hess
eea0a3616c
add thread id field to transferinfo
...
Also converted its timestand to posix seconds, like is used in the other
log files.
2012-07-18 18:42:41 -04:00
Joey Hess
b7d3cefde9
merge two shouldTransfer checks
2012-07-17 12:06:35 -04:00
Joey Hess
cd168c6cba
fix transferrer thread's use of transfer slots and transfer info files
...
Check first if a transfer needs to be done, using the location log only
(for speed), and avoid occupying a slot if not. Always write a transfer
info file, and keep it open throughout the tranfer process.
Now transfers to remotes seem reliable.
2012-07-07 10:50:20 -06:00
Joey Hess
d954a0ce59
fixed close-together transfer race
...
The issue involved forking and they trying to read from a MVar. Reading the
MVar 1st fixed it.
2012-07-06 18:48:51 -06:00
Joey Hess
62876502c5
wait on child transfer processes, and invalidate cache
...
There's still a bug; if the child updates its transfer info file,
then the data from it will superscede the TransferInfo, losing the
info that we should wait on this child.
2012-07-06 16:44:13 -06:00
Joey Hess
4a10795144
logic error
2012-07-06 15:07:42 -06:00
Joey Hess
3d30a45e72
simplified
...
background transferrs seem to work now
2012-07-06 14:54:07 -06:00
Joey Hess
430ad8ce85
it builds again
...
Currently nothing waits on transfer processes.
(Second drive of the day fried. Not concentrating very well.)
2012-07-06 16:41:37 -04:00
Joey Hess
721748135b
fix build (almost)
2012-07-06 14:42:45 -04:00
Joey Hess
a92f5589fc
unfinished (and unbuildable) work toward separate transfer processes
2012-07-05 18:57:06 -06:00
Joey Hess
9eaba58dd9
run transfer with copy of annex state
...
This should have made it run concurrently with other annex actions,
but I'm still seeing it serialize. Perhaps I need to forkProcess?
2012-07-05 16:07:49 -06:00
Joey Hess
2136ee4adb
logic error
2012-07-05 14:45:21 -06:00
Joey Hess
71b5ad8398
wrote transfer thread
...
finally!
2012-07-05 14:34:20 -06:00
