The best CP/M machine ever? was Re: Kaypro II

On 08/24/2012 05:37 PM, Chuck Guzis wrote: Simple, Most of the systems ran without interrupts and did zero timekeeping so while the ap was running it was in control, when the ap returned control CPM was not keeping time either. In essence there was no "idle time" as that concept is only known to multitasking or forground background systems that used blocked IO and task lists to manage what the CPU was doing( rather than spinning in loops). For access time reasons CP/M was biased to seek to home (track000) as directory access was the most frequent and common. The CPU could not do anything during DISK IO as that was a tight IO loop without DMA or other assists. When disk got larger Deblocking was common and that results in caching of at least one physical sector but since ram was considered a scarce commodity a whole track cache was rarely done. The later systems with banked ram and running interrupts would include time outs and other features sorely needed and often omitted in early BIOSes. CP/M by design didn't rule out much so it was often up to the BIOS programmer to make things bullet proof. It was a trivial programming task to make writing the directory sectors twice in two places on the disk where one was outside the usual area. However, with disks under 360K reserving an extra track (typically 4K for a DD disk) was considered costly in terms of the limited disk space and access time during write (typically 6ms per track move and two rotation times (.4S at 300rpm). Same issue for MSdos, though that was FAT and there was a secondary fat. Turns out that 95% of the directory kills were either power up or down, was due to disk in place and door closed. The rest were if the system was flaky and would crash during write or randomly. FYI the directory was the first accessible blocks on the disk, and literally that meant after the system tracks. It was a trivial programming task to read the first data area track (usually the directory was less than 4 blocks) to read those and store them in a reserved file marked RO at the very end of the disk or elsewhere as a recovery tool. Though in most cases the failure was limited to one sector and may not have been a directory sector in use. The directory usually spanned more than one physical sector, storage for that was 32 file names per 1k so even for 512byte sectors only 16 file names were at risk. One had to know the file structure and on disk file structures (skew and sector size) to reconstruct. There were also utilities that would assist in this. Things like smart IDE and SCSI disks and later CF, MMC and SD made all this a non issue as they were smart enough in themselves to power down and avoiding incidental and unintended writes. It was mostly a problem unique to early MFM/RLL and Floppy drives and especially bad for the likes of TRS80 where the drive and system often had their own and independent power supplies as the write line was ACTIVE-LOW so if the CPU/floppy box lost power first it wrote a lot of zeros on the current track of all the drives. Allison