Skip to contentSkip to conversation

hpfall.c improvements, thoughts

PATCH In review Pavel Machek 3 participants 3 messages Mar 25, 2009 – May 20, 2009 2 messages not in the archive

Subsystems (2)
lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]

A message that is not in the archive (sent off-list or not yet received): [email protected]

#2 lore raw

PATCH 1/1 hpfall.c improvements, thoughts

Some conversation I had with Pavel,

imo the example code should be at least barely useable as in not causing harm like I sketched in (1.). Maybe someone else picks that up and improves it later on. Patch at the end, quite sketchy and imature as it doesn't check for errors, but certainly better than before.

Christian

Quoted 22 lines · Pavel Machek

Hi!

Quoted 9 lines

I just hacked and installed the hpfall utility a little bit. By that I thought I should share some ideas with you. (made no patch since this is trivial, but important)

1. hpfall *MUST* mlockall(MCL_CURRENT|MCL_FUTURE); itself! Since the Program sits and waits most of the time it becomes very likely swapped out. If it gets woken up when the laptop drops from the table while it is swapped out it actually triggers harddrive activity!

Quoted 6 lines

2. daemon(0, 0); ... Quick and dirty way to daemonize it

3. Realtime priority:
param.sched_priority = sched_get_priority_max(SCHED_FIFO); sched_setscheduler(0, SCHED_FIFO, &param);
Should give a chance that it has less latency when woken up.

Good ideas. I ommited them, thinking that someone else will do it. Can you submit a patch? (akpm, cc: lk, Eric Piel) I no longer have the hardware, and would prefer not to patch code I can't test.

Quoted 8 lines
Quoted 4 lines

.. another thing is that I still use the old IDE driver, I just changed sda to hda, maybe a bandaid is to use argv[1] for defining the drive (well I understand that it is very basic example code, so maybe someday someone adds a better option parser)

Option parser would indeed be nice.
Pavel

---

Documentation/hwmon/hpfall.c +9 -1
index bbea1cc..f3cde67 100644
@@ -16,6 +16,8 @@
 #include <stdint.h>
 #include <errno.h>
 #include <signal.h>
+#include <sys/mman.h>
+#include <sched.h>
 
 void write_int(char *path, int i)
 {
@@ -63,6 +65,7 @@ void ignore_me(void)
 int main(int argc, char* argv[])
 {
        int fd, ret;
+       struct sched_param param;
 
        fd = open("/dev/freefall", O_RDONLY);
        if (fd < 0) {
@@ -70,7 +73,12 @@ int main(int argc, char* argv[])
                return EXIT_FAILURE;
        }
 
-	signal(SIGALRM, ignore_me);
+       daemon(0,0);
+       param.sched_priority = sched_get_priority_max(SCHED_FIFO);
+       sched_setscheduler(0, SCHED_FIFO, &param);
+       mlockall(MCL_CURRENT|MCL_FUTURE);
+
+       signal(SIGALRM, ignore_me);
 
        for (;;) {
 	       unsigned char count;
#3 lore raw

Hi!

Quoted 6 lines

Some conversation I had with Pavel,

imo the example code should be at least barely useable as in not causing harm like I sketched in (1.). Maybe someone else picks that up and improves it later on. Patch at the end, quite sketchy and imature as it doesn't check for errors, but certainly better than before.

Thanks. Acked-by: Pavel Machek <[email protected]>
Pavel

Quoted 38 lines

---

Documentation/hwmon/hpfall.c +9 -1
index bbea1cc..f3cde67 100644
@@ -16,6 +16,8 @@
 #include <stdint.h>
 #include <errno.h>
 #include <signal.h>
+#include <sys/mman.h>
+#include <sched.h>
 
 void write_int(char *path, int i)
 {
@@ -63,6 +65,7 @@ void ignore_me(void)
 int main(int argc, char* argv[])
 {
        int fd, ret;
+       struct sched_param param;
 
        fd = open("/dev/freefall", O_RDONLY);
        if (fd < 0) {
@@ -70,7 +73,12 @@ int main(int argc, char* argv[])
                return EXIT_FAILURE;
        }
 
-	signal(SIGALRM, ignore_me);
+       daemon(0,0);
+       param.sched_priority = sched_get_priority_max(SCHED_FIFO);
+       sched_setscheduler(0, SCHED_FIFO, &param);
+       mlockall(MCL_CURRENT|MCL_FUTURE);
+
+       signal(SIGALRM, ignore_me);
 
        for (;;) {
 	       unsigned char count;
Signature · 2 lines
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
Acked-by on v1 1/1
#4 lore raw
In reply to Christian Thaeter

Op 25-03-09 22:27, Christian Thaeter schreef:

Quoted 7 lines

Some conversation I had with Pavel,

imo the example code should be at least barely useable as in not causing harm like I sketched in (1.). Maybe someone else picks that up and improves it later on. Patch at the end, quite sketchy and imature as it doesn't check for errors, but certainly better than before.

Hello Christian,
Would you mind resending the patch with a changelog, so that Andrew can merge it?

Thanks,
Eric

Quoted 2 lines
Quoted 37 lines

---

Documentation/hwmon/hpfall.c +9 -1
index bbea1cc..f3cde67 100644
@@ -16,6 +16,8 @@
 #include <stdint.h>
 #include <errno.h>
 #include <signal.h>
+#include <sys/mman.h>
+#include <sched.h>
 
 void write_int(char *path, int i)
 {
@@ -63,6 +65,7 @@ void ignore_me(void)
 int main(int argc, char* argv[])
 {
        int fd, ret;
+       struct sched_param param;
 
        fd = open("/dev/freefall", O_RDONLY);
        if (fd < 0) {
@@ -70,7 +73,12 @@ int main(int argc, char* argv[])
                return EXIT_FAILURE;
        }
 
-	signal(SIGALRM, ignore_me);
+       daemon(0,0);
+       param.sched_priority = sched_get_priority_max(SCHED_FIFO);
+       sched_setscheduler(0, SCHED_FIFO, &param);
+       mlockall(MCL_CURRENT|MCL_FUTURE);
+
+       signal(SIGALRM, ignore_me);
 
        for (;;) {
 	       unsigned char count;