IDE HDD emulation random read/write errors

[Hibou57 (Yannick Duchêne)]

I unfortunately can’t give more tracks other than how to reproduce the bug, especially that the bug occurs randomly.

Basically, I’m trying to install DOS 6.22 on an emulated ISA machine, and it fails, DOS complaining about read or write error on drive C. Repeating the operation multiple time, I see it occurs at random stage, sometime even before it partitions the drive, sometime when it formats the drive, sometime when it copies the files from the floppy to the drive.

To test it, unpack the attached archive and execute `./run` from the extracted directory. The archive contains three raw floppy images for installing DOS 6.22, and a Bourne Shell script which invokes QEmu. Just press enter at any installation stage, the bug may occurs at any stage.

I tried with `cache=none` to be sure it’s not a cache issue, but its the same whatever the cache policy is.

Version and environment: using QEmu 3.0 on Ubuntu 16.04 on a 32 bits DELL Inspiron 9400 (not an emulation, that’s my real laptop).

For why I’m using QEmu for this: the installation proceeds with not error in VirtualBox, but I wanted to use QEmu to have a serial mouse which is not available with QEmu and to have finer control over the machine configuration ; VirtualBox although good, is more limited in that regard.

[Thomas Huth (th-huth)]

The QEMU project is currently considering to move its bug tracking to another system. For this we need to know which bugs are still valid and which could be closed already. Thus we are setting older bugs to "Incomplete" now.
If you still think this bug report here is valid, then please switch the state back to "New" within the next 60 days, otherwise this report will be marked as "Expired". Or mark it as "Fix Released" if the problem has been solved with a newer version of QEMU already. Thank you and sorry for the inconvenience.

[Peter Maydell (pmaydell)]

I tried the repro case five times and each time it ran OK to the point of asking for floppy 2. So that suggests we probably fixed whatever it was...

[Thomas Huth (th-huth)]

Thanks, Peter! Looking at the bug description again and at the point in time when it happened, this was maybe a bug that we also saw with FreeDOS, caused by the CONFIG_ATA_DMA setting in Seabios, which had been fixed here: https://patchwork.kernel<email address hidden>/
Anyway, let's mark this as fixed, if it happens again, then please re-open or file a new bug.