Opened 13 years ago
Closed 13 years ago
#469 closed defect (fixed)
tcp crashes during or after the first HTTP request
| Reported by: | Jakub Jermář | Owned by: | Jiri Svoboda |
|---|---|---|---|
| Priority: | blocker | Milestone: | 0.5.0 |
| Component: | helenos/net/tcp | Version: | mainline |
| Keywords: | Cc: | ||
| Blocker for: | Depends on: | ||
| See also: |
Description
As of mainline,1583, the tcp server crashes when run on amd64 shortly after or during the processing of TCP traffic of the first HTTP request.
The immediate reason for the crash is that in the following snippet:
static void tcp_sock_notify_data(socket_core_t *sock_core)
{
log_msg(LVL_DEBUG, "tcp_sock_notify_data(%d)", sock_core->socket_id);
async_exch_t *exch = async_exchange_begin(sock_core->sess);
sock_core->sess is a non-canonical address, which causes the task to be killed due to a general protection fault. Here, tcp_sock_notify_data() is called from tcp_sock_recv_fibril().
Change History (3)
comment:1 by , 13 years ago
| Summary: | tcp crashes on amd64 → tcp crashes during or after the first HTTP request |
|---|
comment:2 by , 13 years ago
| Priority: | major → blocker |
|---|
Making this a blocker for the 0.5.0 release. We really want to release a functional TCP server.
comment:3 by , 13 years ago
| Resolution: | → fixed |
|---|---|
| Status: | new → closed |
Hopefully fixed in mainline,1584.
Note:
See TracTickets
for help on using tickets.

Just noticed the same crash happens also on ia32. The exception is a standard pagefault due to an unmapped value in
sock_core->sess.