Ipod hard lock while Paused... workaround inside!

[yegods]

not sure if this is the right place, but this problem has plagued me since i started using Rockbox.  the player would freeze up whenever i had the music paused for more than a few seconds, or when using the pause feature and then shutting down.  the player would basically not respond to anything unless i performed the hard reset function.

this patch that i created: http://www.rockbox.org/tracker/task/5987

solves the problem by making the player STOP instead of PAUSE when pausing in the WPS screen.  i've been using it for a week, and have had no freeze-ups in the WPS screen.

the freeze-up issue when paused still needs to be looked at by the devs.  it is a serious issue, and seems to affect most mid-to-late generation ipods.  but in the meantime, this patch will "fix" it.

[GodEater]

Hasn't ever happened to me.

[yegods]

well, it's happened to me, and a lot of other people.  and it's ANNOYING!

[bluebrother]

well, if there is a task in the tracker already -- why is there a need to note it in the forums and linking to the task? Most devs rarely read the forums and so far there weren't much reports of that issue.

[yegods]

well for one, i hadn't received any replies at all to the bug fix, and i thought maybe people read the forums more than they peruse the bug lists.  for two... how can you say there aren't many reports of this?  the second reason i posted was because i saw several reports of this on the front page.

[yegods]

sorry, just thought of a third reason:
there seems to be a misrepresentation that this problem is because of a low or dead battery.  this is totally untrue in a lot of cases.  they seem to have the same symptoms, but i suspect they have different causes.

[GodEater]

Could you please stop double posting ?

[yegods]

sorry man... i didn't know there was a modify function.

but now that you've posted... why are no devs interested in this problem?  i suspect it is affecting more than just Ipods.  there is something crudded up about the Pause functionality.  Removing it has totally gotten rid of any hard locks for me.  this makes my Rockbox experience 200% better at least.

yet i scan through the bug postings and see all these bugs about freezing... most of them related to Pause... seems it would be a bigger issue.  very curious.

[Llorean]

1) Why do you suspect it happens to more than iPods if nobody has reported it for a non-iPod?

2) Where are "all these bugs about freezing" that specifically mention Pausing? I've not noticed a particular trend to Pausing, but if you could cite all the examples, I'll certainly investigate.

Considering I both have an iPod and pause it frequently, I think if we're going to make assumptions, one of the big ones is "it doesn't affect all iPods, so the developers will need one that experiences this problem to investigate it further, unless you're willing to investigate it yourself, because it's very hard to fix something you can't see." And of course, removing the Pause function doesn't *fix* anything.

[yegods]

1) Why do you suspect it happens to more than iPods if nobody has reported it for a non-iPod?

personally, i'm suspecting that no one has put it together, that Pause is causing this problem... that's why no one has reported that specific bug.  however, if you look through the bugs, you'll begin to see there might be a connection... here are some: 7109, 7304, 6846, 6973, 6606, 5987, and on and on, not all involving Ipod.

2) Where are "all these bugs about freezing" that specifically mention Pausing? I've not noticed a particular trend to Pausing, but if you could cite all the examples, I'll certainly investigate.

see above

Considering I both have an iPod and pause it frequently, I think if we're going to make assumptions, one of the big ones is "it doesn't affect all iPods, so the developers will need one that experiences this problem to investigate it further, unless you're willing to investigate it yourself, because it's very hard to fix something you can't see." And of course, removing the Pause function doesn't *fix* anything.

i fairly certain that you're right.  it doesn't seem to affect all ipods, but it certainly does affect more than just mine.  i never said that removing the Pause function fixed it, i said that it "worked around" the problem.  before i created the patch, i would have to hard reset nearly every time i wanted to listen to the ipod.  after i installed the patch, i have only had to hard reset a couple times, and both because i had some other faulty code that i quickly fixed.

sure, the evidence is mostly circumstantial, but it is compelling to me.

thanks for being interested

[Llorean]

7109: Clearly not simply frozen, because he can attempt to start a different song. I don't know your definition of frozen, but we tend to include "does not accept input" and usually "user feedback ceases" as part of it. So clearly his player isn't frozen if he can try to start other songs.
7304: He says pressing play works, which means unpausing works for him, only the power cable insertion causes the freeze.
6846: He can't TURN IT ON. This is a problem that is outside of Rockbox, his Apple firmware is screwing up. Pause isn't even involved, nor is it Rockbox that is freezing.
6973: Again, there is no pausing of playback involved. In this case playback is completely stopped, not Paused.
6606: Again, no Pause involved. Playback is actually *going* when the iPod stops responded.
5987: This one was quite some time old, on UNFINISHED CODE. Your problem is on a different player with almost entirely different hardware, and the ONLY recent posts on it are made by you.

The only problem that MIGHT be related is 5987, and Dan_A, who confirmed it in the past, is one of our 3G developers. I'm quite certain that the problem that bug described had to do with the state of the code for the 3G back then, and is unrelated.

Please, when you say there are "many related tasks" don't try to back it up with tasks that are very different problems, many not involving the pause functionality at all.

I'm more than happy to listen to you explain if I've misread any of these tasks, but at the moment it seems quite clear you're jumping to conclusions, and making broad assumptions.

[yegods]

i said it was circumstantial, didn't i?  all i can go on is what happened to me, and what fixed it.  maybe there are details of what was locking up my ipod that were not made clear.  simple pressing the pause button was only one of the ways that the ipod would regularly lock up.  the other way was that i would shut the ipod down by HOLDing the pause button until it shut down.

this of course led me down the whole "ipod shutdown is messed up in the hardware" patch, which i've pretty much proved is bogus.  once i connected the pause feature to my problem, it was a simple step to stop instead of pause, and test to see if my player would lock up still.  it hasn't locked up after shutdown (meaning i can't turn it on) or pause ONCE since i installed the patch.

now i'm not stupid, and i can put 2 and 2 together, and come up with 4 just like the next guy.  i'm also quite a capable programmer... so it doesn't really matter to me whether anyone looks at the problem, since i've fixed it for myself.  i was just trying to do the rockbox community a favor, and point out what MIGHT be a problem with whatever is going on when the music is paused.

[Llorean]

Yes, but many of those have nothing to do with the Pause button. My point is: Investigate the problem, rather than jumping to conclusions on circumstantial evidence. There are often related problems around here, but I've been working with this project specifically for years, and I'm quite familiar with many things.

For one, you didn't ever SAY holding the Pause button was the problem. Also, your fix would NOT fix holding the Pause button: Holding it never pauses the music, it skips to stop or shutdown depending on how long you hold it, so removing the pause action wouldn't do anything to resolve that if the problem were related to the Pause action. In essence, by adding this information, you've disproved some of your own theories yourself.

Please, don't throw around circumstantial things because it confuses people who don't have the knowledge or ability to research your somewhat wild claims. Instead do the research THEN present the evidence, rather than the other way around.

[yegods]

fair enough, but what i'm suspecting is that there is something in common with how the Pause behaves, and what happens when the shutdown occurs, that causes the lockup.  how else to explain that i never get lockups anymore, even though i use the shutdown all the time now?

maybe i'll dig some more into the firmware, and see if i can find the root cause.  but i just thought that there would be someone more intimately familiar with the playback stuff than i am, and who would be a better one to look at the problem.