ROS testing on real hardware

[roytam]

I don't know what title I should put in jira, and how I should describe these machine specified bugs.

Have you been able to install the revision on any of your computers? At what point is it failing, in the first stage of the installation, second stage or the first time after it is fully installed?

Depending at what point it is failing, I suggest you put something like this in Summary section

"Installing revision 62348 on my VIA PC2500 fails in the first stage with a bsod",

then in the Description section, say what you do and what happens, then say "see via-pc2500.log" or whatever you have called the debug.log.

Up-load your debug log at Attachment, use the browse button to locate the debug.log on your computer.

You will need to put the computer's specification in the Environment section.

These four sections are the only ones you need to fill in.

I hope that this is of some help

I don't install but trying livecd.

[roytam]

another trial with VIA PC2500 system, spec: http://www.via.com.tw/en/initiatives/em ... mainboard/

since this system doesn't boot from USB, I burnt a CD-RW for testing.

and FSE again.

Feel free to add it here too: https://www.reactos.org/wiki/PC_ROS_Rigs

yeah updated.
https://www.reactos.org/wiki/index.php? ... ldid=34435

[Z98]

Wait, livecd? Verify whether the bootcd works or not, livecd regressions aren't tracked in detail by the developers because the livecd is something of a hack.

[erkinalp]

Wait, livecd? livecd is something of a hack.

Is there any plan to have a proper Live CD image on the later stages(capable of installs, network autodetect, independency of builiding environment etc.)?

[Z98]

All the plans in the world don't mean a thing unless someone actually does the work.

[Webunny]

All the plans in the world don't mean a thing unless someone actually does the work.

Lol. Well, I guess there's some truth in that.

Anyway, I'm in the middle of testing the latest build (62368) on RH.

Everything seems to work reasonably well, at least, not worse than one could expect it. I'm happy to note that I don't get a BSOD anymore when I click on 'shut down' or 'restart'.

That said: I still can't drag&drop. Was it meant to be in there? Is that only using the new shell, or should it work with the old shell too?

Edit: asked around on irc, it seems you have to move your mouse-cursor constantly. When you do a 'swipe' while holding it down and dragging and dropping it while moving, it works most of the time. Which means it isn't really stellar. I, or anyone for that matter, that didn't know about the 'moving around', would swear it ain't working.

Another thing: F5 doesn't seem to work on my RH. Right-click refresh did work, though.

Also, I wanted to try out the new explore_new, so I clicked on it (the link/name), but nothing happened. But I heard it's only there in a branch, so I guess it's normal it didn't work.

Lastly, I tried usb 'out of the box' (without any additional patches or trying workarounds out), but it still didn't work.

It was an ok build, I guess, but nothing really special, nor much improvement noticeable from 0.3.16.

[roytam]

Wait, livecd? Verify whether the bootcd works or not, livecd regressions aren't tracked in detail by the developers because the livecd is something of a hack.

because bootcd don't boot as well (and with less log)

[jimtabor]

Wait, livecd? Verify whether the bootcd works or not, livecd regressions aren't tracked in detail by the developers because the livecd is something of a hack.

because bootcd don't boot as well (and with less log)

Bootcd does not help with back tracing too.

[ConductiveDielectric]

No, but for the bug I'm currently experiencing the screen boot mode works really well because driver freezes the system during the boot.

If I may ask: do you get a BSOD too when you log off?

I cant log out from the system because it BSODs during the boot. If you are talking about livecd with removed floppy.sys, I didn't test that use case. Is it somehow related to the floppy driver?

[theuserbl]

Because 0.3.16 don't work on my hardware, I have now tried out the build livecd-62460-dbg.

Running in the debug mode { in the menu at the start: [Reactos (Debug)] }, it ends with a gray screen.

Running in the normal mode { in the menu at the start [ReactOS] }, there comes the error
*** STOP: 0xC000021A (0x0031FDBC,0xC000003C,0x00423C30,0x00000000)

Selecting after [F8] "safe mode" and running then in normal mode { in the menu at the start [ReactOS] }, there comes the error
*** STOP: 0xC000021A (0x0031FDBC,0xC000003C,0x00423C88,0x00000000)

Don't know, why it don't run.
That Windows and Linux working perfect on my computer is obviously.
But Haiku hrev 46642 works on my computer (real hardware), too. ReactOS - on the other side - works not.

Greatings
theuserbl

[Black_Fox]

Running in the debug mode { in the menu at the start: [Reactos (Debug)] }, it ends with a gray screen.

Running in the normal mode { in the menu at the start [ReactOS] }, there comes the error
*** STOP: 0xC000021A (0x0031FDBC,0xC000003C,0x00423C30,0x00000000)

What happens if you start in [ReactOS (Screen)]? Can you get the backtrace when the bugcheck happens in this mode (typing 'bt', but you probably already know)?

[theuserbl]

What happens if you start in [ReactOS (Screen)]?

Code: Select all

(ntoskrnl/ex/init.c:2001) Free non-cache pages: b5de0
(ntoskrnl/mm/ARM3/zeropage.c:50) Free non-cache pages: b601b
(ntoskrnl/mm/section.c:4980) CC failure: c000003c
(dll/ntdll/ldr/ldrapi.c:863) NtCreateSection() failed (Status 0xc000003c)
(lib/rtl/priv.c:443) Retrieving token handle failed (Status c000007c)

*** Fatal System Error: 0xc000021a
                       (0x0031FDBC,0xC000003C,0x004238A8,0x00000000)
[7h"
Entered debugger on embedded INT3 at 0x0008:0x809570a8.
kdb:>

Can you get the backtrace when the bugcheck happens in this mode (typing 'bt', but you probably already know)?

No keyboard input is possible.

Greatings
theuserbl

[fred02]

Can you get the backtrace when the bugcheck happens in this mode (typing 'bt', but you probably already know)?

No keyboard input is possible.

Is it a USB keyboard? Then try enabling legacy USB support in BIOS.

[theuserbl]

Is it a USB keyboard?

Yes.

Then try enabling legacy USB support in BIOS.

There isn't such BIOS-option. Only the possibility to change time and date, boot order and so on.
But I have found an old PS/2 keyboard of mine. And you are are right. With that keyboard I can type there...

Code: Select all

(ntoskrnl/ex/init.c:2001) Free non-cache pages: b5de0
(ntoskrnl/mm/ARM3/zeropage.c:50) Free non-cache pages: b601b
(ntoskrnl/mm/section.c:4980) CC failure: c000003c
(dll/ntdll/ldr/ldrapi.c:863) NtCreateSection() failed (Status 0xc000003c)
(lib/rtl/priv.c:443) Retrieving token handle failed (Status c000007c)

*** Fatal System Error: 0xc000021a
                       (0x0031FDBC,0xC000003C,0x004238A8,0x00000000)
[7h"
Entered debugger on embedded INT3 at 0x0008:0x809570a8.
kdb:> bt
Eip:
<NTOSKRNL.EXE:1570a9 (:0 (RtlpBreakWithStatusInstruction))>
Frames:
<NTOSKRNL.EXE:92414 (ntoskrnl/ke/bug.c:1082 (KeBugCheckWithTf))>
<NTOSKRNL.EXE:928aa (ntoskrnl/ke/bug.c:1407 (KeBugCheckEx))>
<NTOSKRNL.EXE:3a5a2 (ntoskrnl/ex/harderr.c:78 (ExpSystemErrorHandler))>
<NTOSKRNL.EXE:3a6b5 (ntoskrnl/ex/harderr.c:154 (ExpRaiseHardError))>
<NTOSKRNL.EXE:3ac44 (ntoskrnl/ex/harderr.c:632 (NtRaiseHardError))>
<NTOSKRNL.EXE:13fb6d (ntoskrnl/include/internal/arch/../i386/ke.h:687 (KiFastCallEntryHandler))>
<NTOSKRNL.EXE:3ced (:0 (KiFastCallEntry))>
<ntdll.dll:c081>
<smss.exe:626b>
<smss.exe:64fd>
<smss.exe:6f82>
<smss.exe:751a>
<smss.exe:b5ef>
<smss.exe:bd56>
<00000000>
kdb:>

Greatings
theuserbl

[hbelusca]

Hi,
please put your debug log in a report in Jira (together with hardware configuration etc...), otherwise devs won't see it.
Thanks in advance.