tinkered a bit with performance for Cortex flash programming. Mainly make it easier to profile as a start.

git-svn-id: svn://svn.berlios.de/openocd/trunk@1380 b42882b7-edfa-0310-969c-e2dbd0fdcd60
This commit is contained in:
oharboe
2009-02-23 21:26:11 +00:00
parent ba9d608df3
commit 39d80bad9b
2 changed files with 40 additions and 22 deletions

View File

@@ -1700,32 +1700,41 @@ int handle_wait_halt_command(struct command_context_s *cmd_ctx, char *cmd, char
return target_wait_state(target, TARGET_HALTED, ms);
}
/* wait for target state to change. The trick here is to have a low
* latency for short waits and not to suck up all the CPU time
* on longer waits.
*
* After 500ms, keep_alive() is invoked
*/
int target_wait_state(target_t *target, enum target_state state, int ms)
{
int retval;
struct timeval timeout, now;
long long then=0, cur;
int once=1;
gettimeofday(&timeout, NULL);
timeval_add_time(&timeout, 0, ms * 1000);
for (;;)
{
if ((retval=target_poll(target))!=ERROR_OK)
return retval;
keep_alive();
if (target->state == state)
{
break;
}
cur = timeval_ms();
if (once)
{
once=0;
then = timeval_ms();
LOG_DEBUG("waiting for target %s...",
Jim_Nvp_value2name_simple(nvp_target_state,state)->name);
}
gettimeofday(&now, NULL);
if ((now.tv_sec > timeout.tv_sec) || ((now.tv_sec == timeout.tv_sec) && (now.tv_usec >= timeout.tv_usec)))
if (cur-then>500)
{
keep_alive();
}
if ((cur-then)>ms)
{
LOG_ERROR("timed out while waiting for target %s",
Jim_Nvp_value2name_simple(nvp_target_state,state)->name);