OSDev.org

I've recently been doing some Os development on windows and have been using qemu for testing. Having added so gdt code (which is proberbly wrong) I tested as usual. quemu froze and eventualy windows shut down (proberbly a resault of me presing ctrl-alt-del half a million times). I thought the objective of an emulator was to prevent these things happening. Have any of you had similar experiences?

qemu version: 0.8.9-windows

and the code in my os that causes it:


I just tried loading ds with a bad valuse. It had the smae resault. Is this right?

First off, QEMU's latest release is 0.8.2 (Not 0.8.9..)

Secondly QEMU does hang on various specific kernel faults, But I'm sure you could be more specific with your topic.. Are you using Windows 9x? Or NT(2k/xp..).

If QEMU hangs and you don't kill it time, It can constantly use 100% CPU and cause your system to shutdown. (Overheat?)

I use QEMU on OpenBSD and it does crash at times, Usually as result of user error though, But when it does.. It never locks down my entire system

I just need to send SIGKILL and it goes bye-bye

Another reason NOT to use Windows.

(QEMU is also not officially supported on Windows..)

Qemu on windows in in alpha status. Probably for this very reason. Though the 100% CPU usage problem seems to be universal.

I've noticed that the 100% CPU thing is only present when using the SDL interface, Using my patched OpenBSD branch with the curses interface.. It uses around 5%~ when in an idle state.

100% CPU usage shouldn't happen all the time on OpenBSD under SDL (or any *nix I dunno about windows). It should only happen when the guest operating system is using the processer at 100%.

If your guest OS is idling and the host is running at 100% then the guest needs fixing.

Andrew

True, But if it's a closed source guest... Your pretty much out of luck

two things:
1) are you running with kqemu (if so, disable that)
2) are you running a pre-NT style windows (9x, ME) (if so, poor you)

Brynet-Inc, if you have 3 diffrent pieces of software and find bugs in all of them, which is most convenient, to download the source, understand it, debug it, fix it, or to wait for the next version?

I'm just saying that open source software isn't the solution to all problems.

Sure it is bubach, Stop spreading your lies..

You don't like me very much eh? (It's not like your the first..)

Warning of flamewar.

Open source programs are not the best. Maybe you think it because they are free to download and free to modify source (under gnu gpl license). Much programs, that are "open source" are unstable, buggy and un-betatested.

Paid software - well they are not the best too:
Some are very pricy, but they may have quality.
Some are cheap, but the quality is miserable.

So stop scrapping around this, or this whole topic may be locked by moderator, OK?

inflater

Programmers who suck, they write sucky code even if they'd do it for free. There are two kinds of open source code, such that is long and most of it is useless stuff. And another which is short and all useful.

It's the death in open source if your sources are long and full of object spagetti.

In paid software this is not exactly the same. In corporate world, they do not recognize good programmer from a sucky programmer very well so sucky programmers can survive without unsuckifying themselves.

Also what can happen is that managers tend to put sucky programmers together because sucky programmers always underestimate how long it will take them to finish their work. In other words, management likes it when people tell them what they want to hear.

Thats not true, I just think you should keep your thoughts to yourself sometimes..

If everyone kept their thoughts to themselves.. This forum would have no meaningful purpose now would it

Besides, I'm right.. and your wrong.. wanna fight about it~! *shakes fist* j/k

Yeah. Now someone asks, for example, "How to install OpenBSD" ? and someone says: "Type FORMAT C:", do you, Brynet?

inflater