Closed GoogleCodeExporter closed 9 years ago
Seems to be, something strange happen with core.dumped.ru build environment. I
can't reproduce problem on my self-compiled r2324 on WL500gP (I don't own
WL500W).
Can you compile FW yourself? Have you serial console?
Original comment by lly.dev
on 7 Nov 2010 at 3:44
Hi lly,
yes, I have a serial console. But I fear, I'm "too stupid" to compile thе FW
by myself. Or better to say, I never did that before. Will it help, if I'll
look for mesages over the serial interface in the "bricked" state?
Regards
akbor
Original comment by R_K...@gmx.de
on 7 Nov 2010 at 4:16
Serial console should show message of the death. But in case of some problem on
build env. (do you remember that core.dumped.ru is not under our control), this
can be useless.
Anyway, did you try r2314?
Original comment by lly.dev
on 7 Nov 2010 at 4:27
Honestly, I don't know under whose contol is core.dumped.ru :-[
But in any case, I want to help if I can. I didn't try r2313, r2314 and r2322.
What shall I better do? Shall I try one of these releases? Or shall I wait for
the next release and try it?
Original comment by R_K...@gmx.de
on 7 Nov 2010 at 4:40
"Unlimited" from wl500g.info forum create core.dumped.ru resource for public.
Big thanks to him for that.
As you wish. Typical cause of the "brick" (I don't like this word, since
built-in TFTP still work) are in kernel - in your case it _might_ be r2318.
Original comment by lly.dev
on 7 Nov 2010 at 4:53
OK, then I'll wait for the next available release on core.dumped.ru an try it.
CU
Original comment by R_K...@gmx.de
on 7 Nov 2010 at 4:58
The only patches from r2318 commit which can break normal boot is:
164-timerfd1-newapi.patch
164-timerfd2-fs.patch
but I still can't get stable reproduction of problem/bug.
Original comment by lly.dev
on 9 Nov 2010 at 7:43
today I've flashed the new r2326 (WL500W-1.9.2.7-rtn-r2326.trx) with the same
result: device doesn't come up, all ETH activity LEDS blink.
Serial console repeats the same output cyclically. I've logged two cycles, see
attached text file
reagrds
Original comment by R_K...@gmx.de
on 9 Nov 2010 at 9:00
Attachments:
Some kind of induced problem :(
Can you attach console log of successful boot on your 500W?
Original comment by lly.dev
on 10 Nov 2010 at 7:32
OK, I will give you a normal boot log today evening. I've already restored my
device with r2308 again. I could also try to find out, which is the first
release with this bug (between r2308 and r2324), but I need a little bit more
time for that.
Original comment by R_K...@gmx.de
on 10 Nov 2010 at 9:51
Problem appeared on MIPS32R1 devices only (WL500W, WL500gP) and IMHO caused not
by some commit text itself. Problem either in serial initialization code for an
old devices (RT-N16 works well), or it is gcc optimizer bug.
Original comment by lly.dev
on 10 Nov 2010 at 9:59
I have the similar problem with WL500W-1.9.2.7-rtn-r2326.
I have attached the log file which captured from console with some cycles of
booting. I hope it can be helpful. Thanks.
Original comment by sergei.v...@gmail.com
on 10 Nov 2010 at 11:36
Attachments:
Rollback to WL500W-1.9.2.7-rtn-r2313.
Attached console log of successful boot. Thanks.
Original comment by sergei.v...@gmail.com
on 10 Nov 2010 at 7:05
Attachments:
OK, just in this moment I hold a screwdriver in the hand and want to open my
WL-500W and to get the successiful boot log. Goot timing, then I don't need to
do that :)
Original comment by R_K...@gmx.de
on 10 Nov 2010 at 7:25
Hi,
today I tried all releases through. r2314 is the last working release, from
r2322 and higher my WL-500W does cyclically reboot. Maybe that will help to
lokalize the bug.
regards
Original comment by R_K...@gmx.de
on 12 Nov 2010 at 2:00
Unfortunately, problem reproducible not with 100% and even debugging printk's
often hide the bug. r2318 isn't buggy, it only uncover an old bug.
Original comment by lly.dev
on 12 Nov 2010 at 4:33
I tried to fix problem in r2332 (should fix some potential races), but it might
be additional patches needed.
Original comment by lly.dev
on 12 Nov 2010 at 5:39
How can I test it? Should I wait on the next release on core.dumped.ru?
Original comment by R_K...@gmx.de
on 12 Nov 2010 at 7:41
[deleted comment]
As usual, you have at least two choices:
1) compile FW yourself
2) use night build from core.dumped.ru
Original comment by lly.dev
on 13 Nov 2010 at 7:03
OK, I'll wait for the next build.
Original comment by R_K...@gmx.de
on 13 Nov 2010 at 7:20
r2334 already on core.dumped.ru.
Anybody yet was tried it on WL-500W?
Original comment by sergei.v...@gmail.com
on 13 Nov 2010 at 8:04
I tried r2334 just now. Seems to work! lly, thank you very much!
Original comment by R_K...@gmx.de
on 13 Nov 2010 at 10:08
In case of similar problems just reopen this issue.
Original comment by lly.dev
on 14 Nov 2010 at 10:51
Original issue reported on code.google.com by
R_K...@gmx.de
on 7 Nov 2010 at 11:17