Opened 15 years ago
Closed 15 years ago
#206 closed defect (fixed)
nettest1 hangs on sparc64
Reported by: | Jakub Jermář | Owned by: | |
---|---|---|---|
Priority: | major | Milestone: | 0.4.3 |
Component: | helenos/net/other | Version: | mainline |
Keywords: | Cc: | ||
Blocker for: | Depends on: | ||
See also: |
Description
When I run nettest1 on HelenOS/sparc64, the test hangs after the first "Sendto and recvfrom" is printed. The system's performance seems to be severely impacted even though the system continues to slowly react.
Note:
See TracTickets
for help on using tickets.
In the first invocation of sendto_core(), socket→header_size is plainly too big: 34359738368. This results in number of
fragments
computed as 524456. The system then performs this many calls to
async_data_write_start()
.
So the question is why is socket→header_size thus big on sparc64 and where is this value set.