Virtual PC with ReactOS 0.3.8 : ok
Moderator: Moderator Team
Virtual PC with ReactOS 0.3.8 : ok
Hi everybody,
Just an interesting info for who'd like to test : ReactOS 0.3.8 now works with Microsoft Virtual PC.
The keyboard driver works well ; the mouse driver, however, seems to be crippled (pointer moves but with some lag).
Tarnyko
Just an interesting info for who'd like to test : ReactOS 0.3.8 now works with Microsoft Virtual PC.
The keyboard driver works well ; the mouse driver, however, seems to be crippled (pointer moves but with some lag).
Tarnyko
Re: Virtual PC with ReactOS 0.3.8 : ok
That's great news!
Re: Virtual PC with ReactOS 0.3.8 : ok
We were aware of new VPC version finally working with ROS. Good to see it confirmed.
Re: Virtual PC with ReactOS 0.3.8 : ok
bad news...
i hav just tested r39838 and it doesnt boot under MS Virtual PC....
i hav just tested r39838 and it doesnt boot under MS Virtual PC....
Re: Virtual PC with ReactOS 0.3.8 : ok
for more infomation, i get this message:
"partition signature != 55AA"
1st stage finishes successfully, but this message is displayed at 2nd stage boot.
"partition signature != 55AA"
1st stage finishes successfully, but this message is displayed at 2nd stage boot.
Re: Virtual PC with ReactOS 0.3.8 : ok
The ReactOS 0.3.10 does not work in Virtual PC 2007.
In time to install the screen freezes after a few seconds...
Is there a way to solve the problem?
Sorry my English.
In time to install the screen freezes after a few seconds...
Is there a way to solve the problem?
Sorry my English.
Re: Virtual PC with ReactOS 0.3.8 : ok
Microsoft better patch this right away! This just isn't like them at all!Hi everybody,
Just an interesting info for who'd like to test : ReactOS 0.3.8 now works with Microsoft Virtual PC.
The keyboard driver works well ; the mouse driver, however, seems to be crippled (pointer moves but with some lag).
Tarnyko
-
- Test Team
- Posts: 802
- Joined: Thu Apr 03, 2008 2:17 pm
- Contact:
Re: Virtual PC with ReactOS 0.3.8 : ok
This isn't a fault with Microsoft; Virtual PC works fine with Windows. There's a problem with our drivers, notably the serial port driver.Ubuntu wrote:Microsoft better patch this right away! This just isn't like them at all!Hi everybody,
Just an interesting info for who'd like to test : ReactOS 0.3.8 now works with Microsoft Virtual PC.
The keyboard driver works well ; the mouse driver, however, seems to be crippled (pointer moves but with some lag).
Tarnyko
Re: Virtual PC with ReactOS 0.3.8 : ok
Do you have any estimates of when going to work in Virtual PC 2007?Lone_Rifle wrote:This isn't a fault with Microsoft; Virtual PC works fine with Windows. There's a problem with our drivers, notably the serial port driver.Ubuntu wrote:Microsoft better patch this right away! This just isn't like them at all!Hi everybody,
Just an interesting info for who'd like to test : ReactOS 0.3.8 now works with Microsoft Virtual PC.
The keyboard driver works well ; the mouse driver, however, seems to be crippled (pointer moves but with some lag).
Tarnyko
Does this link help on something?
http://www.microsoft.com/windows/produc ... debug.mspx
It has a. Doc describing the debug of VPC 2004, may perhaps serve to VPC 2007?
Re: Virtual PC with ReactOS 0.3.8 : ok
My quote was sarcasm. Microsoft has a history of blocking the competition.This isn't a fault with Microsoft; Virtual PC works fine with Windows. There's a problem with our drivers, notably the serial port driver.
Re: Virtual PC with ReactOS 0.3.8 : ok
I am trying to play with this some to figure out what the issue with the mouse is. What I would like is to get the various logging things in the i8042 driver to be written somewhere - ideally to the COM port where I can see it in putty. But so far I get no output..
I hacked i8042prt.c to have this:
#if 1
DbgSetDebugFilterState(
DPFLTR_I8042PRT_ID,
(1 << DPFLTR_ERROR_LEVEL) | (1 << DPFLTR_WARNING_LEVEL) |
(1 << DPFLTR_TRACE_LEVEL) | (1 << DPFLTR_INFO_LEVEL) | DPFLTR_MASK,
TRUE);
#endif
and then rebuilt the cd image, but I still get no output. On the other hand in a cmd window I can do "dir > COM2:" which causes display to putty (there is a weird off-by-one issue here - putty thinks it is listening to the pipe associated with COM1).
which on the surface would seem to do what I want. I guess the thing I don't really have a good sense about is where one would expect the debug output to really go - I am guessing that it is supposed to go to a COM port.
Can someone point me in the right direction?
I hacked i8042prt.c to have this:
#if 1
DbgSetDebugFilterState(
DPFLTR_I8042PRT_ID,
(1 << DPFLTR_ERROR_LEVEL) | (1 << DPFLTR_WARNING_LEVEL) |
(1 << DPFLTR_TRACE_LEVEL) | (1 << DPFLTR_INFO_LEVEL) | DPFLTR_MASK,
TRUE);
#endif
and then rebuilt the cd image, but I still get no output. On the other hand in a cmd window I can do "dir > COM2:" which causes display to putty (there is a weird off-by-one issue here - putty thinks it is listening to the pipe associated with COM1).
which on the surface would seem to do what I want. I guess the thing I don't really have a good sense about is where one would expect the debug output to really go - I am guessing that it is supposed to go to a COM port.
Can someone point me in the right direction?
-
- Test Team
- Posts: 802
- Joined: Thu Apr 03, 2008 2:17 pm
- Contact:
Re: Virtual PC with ReactOS 0.3.8 : ok
We have some known issues with i8042prt, although the problems haven't actually been attended to. The relevant bug report for this would be
http://www.reactos.org/bugzilla/show_bug.cgi?id=701
http://www.reactos.org/bugzilla/show_bug.cgi?id=701
Re: Virtual PC with ReactOS 0.3.8 : ok
Possible, I guess. The problem I see on VirtualPC is that the mouse moves in a very jerky fashion - almost as if it is polling the device every second or two instead of responding immediately to interrupts. And for that matter, double-clicking doesn't work on VirtualPC.
The queueing issue in the bug report could well be it, I suppose.
The queueing issue in the bug report could well be it, I suppose.
Re: Virtual PC with ReactOS 0.3.8 : ok
Update: I have spent some time fooling with this, and there are some rather strange things going on under VirtualPC. Essentially I can't get anything working at all that involves COM ports and the kernel using VirtualPC. No GDB, no KDB, no DebugPrint messages. If I attempt to use Tab-K, the system freezes as if it is attempting to bring up the debugger, but I get no output.
From a cmd window, I can do "dir > COM2" and the output comes to COM1 (notice the off-by-one error in the naming).
I could dump the DebugPrint messages to a file which sort of told me how the mouse was initialized, but at some point after the mouse is initialized the log was truncated and there were no further messages. The mouse does move in a sort of jerky fashion, but there are no more messages.
As a sanity check, I ran a Sun VirtualBox, and everything behaves normally. I get the debug output, Tab-K brings up the kdb> prompt. Everything.
All of this together (the odd mouse behavior, the COM port problems, and the inability to get logging information past a certain point) lead me to suspect that the problem has nothing directly to do with the i8042 driver. For all I know, it might have nothing to do with the serial port either. At the moment I have half-baked ideas that that it relates to irq levels or perhaps secondary threads. I would note however that VirtualPC shows 0% CPU utilization when the VM is idle.
The trick of course is to figure out where this is happening. Which driver, or which part of the kernel? I can disable a number of the drivers that aren't needed for VPC to see if I can narrow down the search somewhat. I already have a "Tab-m" that I added which creates a pagefile and then tries to do a memory dump to it, but this merely seems to wedge the thing without writing anything useful - probably caught in a spinlock somewhere, so instead I am fixing it to dump a threadlist to see what that tells me.
From a cmd window, I can do "dir > COM2" and the output comes to COM1 (notice the off-by-one error in the naming).
I could dump the DebugPrint messages to a file which sort of told me how the mouse was initialized, but at some point after the mouse is initialized the log was truncated and there were no further messages. The mouse does move in a sort of jerky fashion, but there are no more messages.
As a sanity check, I ran a Sun VirtualBox, and everything behaves normally. I get the debug output, Tab-K brings up the kdb> prompt. Everything.
All of this together (the odd mouse behavior, the COM port problems, and the inability to get logging information past a certain point) lead me to suspect that the problem has nothing directly to do with the i8042 driver. For all I know, it might have nothing to do with the serial port either. At the moment I have half-baked ideas that that it relates to irq levels or perhaps secondary threads. I would note however that VirtualPC shows 0% CPU utilization when the VM is idle.
The trick of course is to figure out where this is happening. Which driver, or which part of the kernel? I can disable a number of the drivers that aren't needed for VPC to see if I can narrow down the search somewhat. I already have a "Tab-m" that I added which creates a pagefile and then tries to do a memory dump to it, but this merely seems to wedge the thing without writing anything useful - probably caught in a spinlock somewhere, so instead I am fixing it to dump a threadlist to see what that tells me.
Who is online
Users browsing this forum: No registered users and 50 guests