This is an automated email from the git hooks/post-receive script.
git pushed a commit to branch master
in repository enlightenment.
View the commit online.
commit f5fc3c18db6467238f03098becfa03217541c728
Author: Carsten Haitzler <[email protected]>
AuthorDate: Wed Jan 14 14:47:47 2026 +0000
cpufreq - when in freeze eg scr off go slow in poll eg 60 sec
---
src/modules/cpufreq/cpf.c | 51 ++++++++++++++++++++++++++++++++++++++--
src/modules/cpufreq/cpf.h | 3 +++
src/modules/cpufreq/e_mod_main.c | 9 +++++++
3 files changed, 61 insertions(+), 2 deletions(-)
diff --git a/src/modules/cpufreq/cpf.c b/src/modules/cpufreq/cpf.c
index 768f32e72..3d89056e8 100644
--- a/src/modules/cpufreq/cpf.c
+++ b/src/modules/cpufreq/cpf.c
@@ -28,6 +28,10 @@ static Cpf_Stats *_cpu_stats = NULL;
// current cpu level
static int _cpu_level = -1;
+// wake up cpf poller thread
+static Ecore_Pipe *_cpf_pipe = NULL;
+static int _cpf_pipe_fd = -1;
+
typedef struct
{
Cpf_Event event;
@@ -140,6 +144,11 @@ _thread_main(void *data EINA_UNUSED, Ecore_Thread *eth)
int i, y, pending;
Cpf_Stats *cpf_stat;
Cpf_Render *r;
+ int ret;
+ fd_set rfds, wfds, exfds;
+ struct timeval tv;
+ char buf[1] = { 0 };
+ double t;
// lock we can wait on until this func is done
eina_lock_take(&_poll_thread_lock);
@@ -242,7 +251,25 @@ skip:
eina_lock_take(&_poll_lock);
tim = _poll_time;
eina_lock_release(&_poll_lock);
- usleep(tim * 1000000);
+
+ t = 0.0;
+ while (t < 0.1)
+ {
+ FD_ZERO(&rfds);
+ FD_ZERO(&wfds);
+ FD_ZERO(&exfds);
+ FD_SET(_cpf_pipe_fd, &rfds);
+ tv.tv_sec = tim;
+ tv.tv_usec = (unsigned int)(tim * 1000000.0) % 1000000;
+ t = ecore_time_get(); // how long did we spend
+ ret = select(_cpf_pipe_fd + 1, &rfds, &wfds, &exfds, &tv);
+ t = ecore_time_get() - t; // time spent
+ if ((ret == 1) && (FD_ISSET(_cpf_pipe_fd, &rfds)))
+ {
+ if (read(_cpf_pipe_fd, buf, 1) < 0)
+ fprintf(stderr, "%s: ERROR READING FROM FD\n", __func__);
+ }
+ }
}
eina_lock_release(&_poll_thread_lock);
cpu_perf_del(cp);
@@ -344,6 +371,11 @@ _cb_system_pwr_get(void *data EINA_UNUSED, const char *params)
}
}
+static void
+_cb_pipe_dummy(void *data EINA_UNUSED, void *buffer EINA_UNUSED, unsigned int bytes EINA_UNUSED)
+{
+}
+
void
cpf_init(void)
{ // set everything up
@@ -363,6 +395,9 @@ cpf_init(void)
eina_lock_new(&_poll_thread_lock);
eina_lock_new(&_renders_lock);
eina_lock_new(&_cpu_stats_lock);
+ _cpf_pipe = ecore_pipe_add(_cb_pipe_dummy, NULL);
+ if (_cpf_pipe) ecore_pipe_freeze(_cpf_pipe);
+ if (_cpf_pipe) _cpf_pipe_fd = ecore_pipe_read_fd(_cpf_pipe);
_poll_thread = ecore_thread_feedback_run
(_thread_main, _thread_notify, _thread_end,
_cancel, NULL, EINA_TRUE);
@@ -379,6 +414,7 @@ cpf_shutdown(void)
ecore_thread_cancel(_poll_thread);
_poll_thread = NULL;
eina_lock_take(&_poll_thread_lock);
+ cpf_wake();
eina_lock_release(&_poll_thread_lock);
eina_lock_free(&_poll_thread_lock);
eina_lock_free(&_poll_lock);
@@ -386,6 +422,9 @@ cpf_shutdown(void)
eina_lock_free(&_cpu_stats_lock);
evas_object_del(_renders_win);
if (_cpu_stats) _cpf_stats_free(_cpu_stats);
+ if (_cpf_pipe) ecore_pipe_del(_cpf_pipe);
+ _cpf_pipe = NULL;
+ _cpf_pipe_fd = -1;
_cpu_stats = NULL;
_renders_win = NULL;
_renders_data_cpu_usage = NULL;
@@ -450,7 +489,7 @@ cpf_event_callback_del(Cpf_Event event, void (*cb) (void *data), void *data)
_callbacks = eina_list_remove_list(_callbacks, l);
free(c);
break;
- }
+ }
}
}
}
@@ -549,3 +588,11 @@ cpf_render_unreq(Cpf_Render_Type type, int w, int h)
}
eina_lock_release(&_renders_lock);
}
+
+void
+cpf_wake(void)
+{
+ char buf[1] = { 1 };
+
+ if (_cpf_pipe) ecore_pipe_write(_cpf_pipe, buf, 1);
+}
diff --git a/src/modules/cpufreq/cpf.h b/src/modules/cpufreq/cpf.h
index d79317cd3..3b8c0009c 100644
--- a/src/modules/cpufreq/cpf.h
+++ b/src/modules/cpufreq/cpf.h
@@ -44,6 +44,8 @@ typedef struct
Cpf_Render **rend;
} Cpf_Stats;
+extern Ecore_Pipe *cpf_pipe;
+
void cpf_init(void);
void cpf_shutdown(void);
void cpf_poll_time_set(double tim);
@@ -55,5 +57,6 @@ int cpf_perf_level_get(void); // 0-100
const Cpf_Stats *cpf_perf_stats_get(void); // per core stat/freq info
void cpf_render_req(Cpf_Render_Type type, int w, int h);
void cpf_render_unreq(Cpf_Render_Type type, int w, int h);
+void cpf_wake(void);
#endif
diff --git a/src/modules/cpufreq/e_mod_main.c b/src/modules/cpufreq/e_mod_main.c
index 816b8024d..d528972a2 100644
--- a/src/modules/cpufreq/e_mod_main.c
+++ b/src/modules/cpufreq/e_mod_main.c
@@ -498,6 +498,9 @@ _handle_powersave_mode(E_Powersave_Mode mode)
case E_POWERSAVE_MODE_LOW:
printf("PWSV: low=%i\n", cpufreq_config->power_hi);
cpf_perf_level_set(cpufreq_config->power_hi);
+ // ensure poll time is restored in case we were in freeze
+ cpf_poll_time_set(cpufreq_config->check_interval);
+ cpf_wake();
break;
case E_POWERSAVE_MODE_MEDIUM:
@@ -509,11 +512,17 @@ _handle_powersave_mode(E_Powersave_Mode mode)
case E_POWERSAVE_MODE_EXTREME:
printf("PWSV: extreme=%i\n", cpufreq_config->power_lo);
cpf_perf_level_set(cpufreq_config->power_lo);
+ // ensure poll time is restored in case we were in freeze
+ cpf_poll_time_set(cpufreq_config->check_interval);
+ cpf_wake();
break;
case E_POWERSAVE_MODE_FREEZE:
printf("PWSV: freeze\n");
cpf_perf_level_set(0);
+ // when in freeze - eg screen off - only update every minute
+ cpf_poll_time_set(60.0);
+ cpf_wake();
break;
default:
--
To stop receiving notification emails like this one, please contact
the administrator of this repository.
