Login
Username:

Password:

Remember me



Lost Password?

Register now!
Main Menu
Who's Online
1 user(s) are online (1 user(s) are browsing Forum)

Members: 0
Guests: 1

more...
   All Posts (SonicAmiga)


(1) 2 3 4 ... 48 »


Re: Aros Subpartition and SFS for Linux 3.0.x & 3.2.x

Joined:
2005/4/1 8:40
Group:
Member
Posts: 476
Offline
Just stumbled upon this thread. It's great to see that SFS as adapted to 3.2. I already thought i'll have to learn VFS differences and do this myself.
Unfortunately Linux upstream maintainers don't want to take SFS. I remember that last time when i posted it to LKML they told me: "We are not going to include every obscure filesystem". :(
I think it would be much better to have a separate source archive which can be built as module. I'll take care about it.

Posted on: 2013/7/27 9:45
Transfer the post to other applications Transfer


Re: Windows-hosted AROS - testing needed

Joined:
2005/4/1 8:40
Group:
Member
Posts: 476
Offline
If anyone wishes to backport to v0, the actual changes are:
1. Added 'volatile' to virtual CPU state variables. Fixes compiler's optimization of "while (Sleep_Mode);" to "if (Sleep_Mode) while(1);"
2. SetThreadContext()/ResumeThread() - note changed logic, no more sleeping with disabled interrupts.
These two comprised the bug. It was very hard to find because attempts to change (1) or (2) separately did not work, and this was very misleading. My debug NMI implementation (see tests/windebug.c) together with gdb helped me to shed some light on this mystery.

P.S. I wonder if the concept of debug NMI could be generalized to all platforms? For example i easily see how this could be used on Linux-hosted (by hooking up to SIGHUP for example). But what about real hardware? Do modern machines have any kind of magic buttons/keys/pins/wires/whatever ? Ability to talk to totally dead system and figure out what happened is very attractive.

Posted on: 2012/9/7 6:48
Transfer the post to other applications Transfer


Windows-hosted AROS - testing needed

Joined:
2005/4/1 8:40
Group:
Member
Posts: 476
Offline
Hello!
If anyone remembers, there was a problem with Windows-hosted. It may lock up randomly.
Good news - yesterday i found out what was wrong. And fixed it. Now intensive testing is needed.
At least, works for me now.

Posted on: 2012/9/6 7:20
Transfer the post to other applications Transfer


Re: Android-hosted AROS

Joined:
2005/4/1 8:40
Group:
Member
Posts: 476
Offline
Hi all!
No, your Android does not need to be rooted.
When i last time checked this version, it worked. But now i can't maintain this port because i have no devices capable of running Android.
From the log it looks like there's some problem in communication of AROS display driver with Android part. May me you already had older version installed and did not update the bootstrap? Actually bootstrap and display driver (which is in the BSP) versions should match. If they differ, various things can happen.

Posted on: 2012/9/6 7:16
Transfer the post to other applications Transfer


Re: AROS MinGW Hosted unstable on Windows 7 64-bit

Joined:
2005/4/1 8:40
Group:
Member
Posts: 476
Offline
Lockup is old well-known problem. On some machines it triggers, on some it doesn't. Extremely hard-to-find thing.

Posted on: 2012/8/14 4:12
Transfer the post to other applications Transfer


Anyone here from South Korea ?

Joined:
2005/4/1 8:40
Group:
Member
Posts: 476
Offline
Hi people!
Is there anyone here from South Korea? I'm currently on business trip to Samsung HQ until 15th of September, and i live in Suwon in a hotel. Just want to know, whether i'm alone Amigoid and AROSian here or not?
2 all others: at home i currently don't have Internet connection. Currently i'm using hotel's free wifi.

Posted on: 2012/8/13 7:02

Edited by SonicAmiga on 2012/8/13 7:34:34
Transfer the post to other applications Transfer


Re: Request for testing: ACPI Phase 1

Joined:
2005/4/1 8:40
Group:
Member
Posts: 476
Offline
Quote:

DizzyOfCRN wrote:
I booted with success with a build that was from the repo some hour ago,

Good, it works then.
[quote}
How about adding the svn revision on the "AROS - The AROS Research OS. Compiled Nov 14 2011" line too?
[/quote]
See no problem with this.
[quote}
Seems that grub decided to act differently with the i386 pkg on my system
[/quote]
No, it's AROS bootstrap acting differently. PKG is better than ELF linking because all original module names are preserved. This makes it easier to replace modules, and crash backtrace will show them instead of "Kickstart ELF".

Posted on: 2011/11/15 9:59
Transfer the post to other applications Transfer


Re: Request for testing: ACPI Phase 1

Joined:
2005/4/1 8:40
Group:
Member
Posts: 476
Offline
Just committed the fix. Should work now. However tested only on Mac so far.

Posted on: 2011/11/15 2:41
Transfer the post to other applications Transfer


Re: Request for testing: ACPI Phase 1

Joined:
2005/4/1 8:40
Group:
Member
Posts: 476
Offline
DizzyOfCRN: Thanks, this will help me to fix it. If you look carefully at the log, you'll see overlapping addresses (kickstart destroys boot info).
Currently the bootstrap assumes that boot info won't be put by GRUB after all modules, and on your system this happens to be not true. I'll adapt the bootstrap to handle it.

Posted on: 2011/11/15 1:26
Transfer the post to other applications Transfer


Re: Request for testing: ACPI Phase 1

Joined:
2005/4/1 8:40
Group:
Member
Posts: 476
Offline
QEmu?
I need the complete log up to this point.

Posted on: 2011/11/14 9:22
Transfer the post to other applications Transfer



 Top
(1) 2 3 4 ... 48 »




Search
Top Posters
1 paolone
paolone
3794
2 nikolaos
nikolaos
3470
3 magorium
magorium
3167
4 phoenixkonsole
phoenixkonsole
3091
5 deadwood
deadwood
2387
6 ncafferkey
ncafferkey
2151
7 mazze
mazze
2071
8 clusteruk
clusteruk
2051
9 Kalamatee
Kalamatee
1796
10 damocles
damocles
1772
© 2004-2014 AROS Exec