It's an obscure 1995 magazine CD from my country. I've been dumping everything in my possession lately. The ISO is 134MB bzipped but just running the .exe file without any data is enough to trigger the crash. I believe there would be no problem in distributing this specific executable for you to test.

The crash upon entering the BIOS seems to be fixed, but the one from the log I just posted still happens. I'm guessing it's completely unrelated to the cd subsystem, sorry!

Here is a log from one of these CD-ROM crashes (this time it happened for the first time when trying to execute a program inside the CD, not when accessing the CMOS Setup) New CD command 00 116301737 INT out of range INT out of range INT out of range New CD command 28 116304307 INT out of range INT ...

Here is a new patch:

-Fixes CGA/MDA
-Turbo selection works (fix also applied to the Toshiba t3100/t1000) - Sarah, this is very intrusive to the cpu subsystem and I don't know much about it, please review carefully
-Added option to select flash size and turbo behavior, just like the real hardware

Committed at rev 1073. I fixed access to high density floppies, DOS 6.22 seems to boot for me now. Oops, I forgot to check the "AT" variable in the code. I only checked the MODEL_AT and MODEL_PS2 flags for correct behavior. I can't get CGA, MDA, or Hercules to work. It just doesn't render...

I start the machine it post but when it goes to boot it crashes. It seems to happen when certain cdrom images are mounted at boot. I can confirm that this happened to me some time ago with the install CD of a linux distro, mounted at POST = crash in various AMI bioses for Pentium. Forgot the report.

This is a bug in the latest version, I guess Sergey always uses XTIDE and didn't notice it. I've identified the offending commit and notified him about it in github. I've fixed it in the attached bios. Still it doesn't seem to boot the MS-DOS 6.22 install disk 1, I've only succesfully booted the 5.0...

I use the shaders from leilei's repository with default settings: https://github.com/leilei-/someshaders This patch has a problem: the turbo key combination doesn't work, the CPU always stays a full speed. (under my window manager under linux, I've to use the ctrl-alt-keypad- and ctrl-alt-keypad+ tu...

Just a quick note: under Linux, alt+tab already works fine.

Screenshots

The Xi8088 is a great project by Sergey Kiselev to build a XT-clone, I've built it myself along with the other boards by Sergey and they work great. The system is a XT along with some AT and PS2 features like PS/2 mouse, AT RTC/NVR, two PICs, etc. More info in http://www.malinov.com/Home/sergeys-pro...

I couldn't locate the exact bios rom the patch needs, but the zappa upgrade for Gateway seems to work with an odd FDC issue. Just throwing this screenshot out there since amibios oem logos are cute and stirring interest into more possible OEM model featurecreep What is the FDC issue? This Gateway u...

This issue has bitten me twice, first with real hardware and now with PCem. It is a silent failure. The XTide and the NE2000 addresses are both 0x300, causing a conflict. In real hardware, I've had to put the card into another (more recent, with IDE) computer and run the NE2000 configuration utility...

I must correct myself, I've only tested with Windows 95. The GD5434 Win3.1 drivers are not working, even with today's MMIO fixes. The attached screenshot is from the Win31 Program Manager with a menu open.

While the Oti-037 has fewer capabilities than the already emulated Oti-067, its BIOS is somewhat pickier and did some tests that PCem didn't pass yet with the Oti-067 code, including enabling/disabling the controller, color diagnostic reading and reading from undocumented registers. The diagnostic t...

With the commits in the past week, PCem now emulates every system (sans some vgas and sound cards) that I've had contact prior to ~1996, this is wonderful. The PB520r is working perfectly with my HDD & BIOS dumps. EDIT: Just a quick note: most GD5434 were just 1MB with expansion slots for 2MB (t...

Thanks, Sarah! Sorry for that bit, won't do it again.

By the way, I don't know what's different between the Cirrus 5429/5430/5434 other than bus width and other stuff that should affect only speed, but the 5429 code worked fine with the BIOS files I've extracted from two of my cards (two 5434 and one 5430 - one of them is onboard a Intel Premiere/PCI o...

I've seen it happen in real hardware, but not with the AWE32. Wrong and/or missing music and even sounds in Duke3D under Windows. Generally a reboot would fix it.

Doing test releases before (for what was released as 13.0, for example) is also a good idea to warn users that there are a lot of new features and a bugfix release may be needed. Even with all the testers, some small bugs always pass (like what happened now).

He really meant one port/two drives. I have machines like that.

This PC87306 code was very obviously taken from the 86Box Project. Yes, I said it expressively in my post, can you read? Battler himself told me in the past that 86box had this code. This is fine, but please leave file headers intact and credit the projects and/or authors while posting such patches...

This patch adds support for the Intel Advanced/ZP (Zappa) motherboard for Socket 5. It is similar to the Advanced/EV (Endeavor) board, but has a few nice differences like the BIOS supporting CD-ROM booting (only in the First Boot Device position) and having no non-emulated on-board video/audio clutt...

Mine is 12.5MHz too, that's the reason I made this suggestion. But it's okay, just a small change that I can maintain locally

To anyone having problems with this on Linux: remember about case-sensitivity. Some older cd ripping programs write the cuesheets contents in upper case, including the filename of the .bin. Just edit the cuesheet/rename the .bin and it should work. Or you can use all uppercases when ripping. I can c...

According to every datasheet I could find, the all the -12 80286s were actually 12.5MHz. This is a small difference from the round 12Mhz value currently in the code, but I thought I would report it anyway.

Trying to install Windows 95 with the latest development version results in a bluescreen followed by a freeze at the very end of the installation process (right before beginning the printer configuration step). Invalid segment type for 0A6C! 0000 onesec onesec onesec onesec onesec onesec onesec ones...

The screen is not redrawn when switching virtual desktops on Linux. This is worse if emulating a system without mouse, because then we can't just move the mouse to force the screen to be redrawn.

I've built this card (along with all the other cards by him.) The BIOS is for computers which do not know about the enhanced features present in newer floppy disk controllers, so all you should do is load the BIOS and probably call some functions to set the FDC capabilities in the model you want (se...

That shouldn't be a problem. The software renderer is missing a few features though such as linear filtering, but maybe a message box could pop up the first time explaining that it's there for compatibility reasons and should be changed to OpenGL if possible. But linear filtering should be disabled...