[TUHS] DG UNIX History
bakul at iitbombay.org
Wed Nov 16 01:48:23 AEST 2022
On Nov 15, 2022, at 7:11 AM, Larry McVoy <lm at mcvoy.com> wrote:
> On Tue, Nov 15, 2022 at 01:03:05AM -0500, Theodore Ts'o wrote:
>> P.S. I remember back in the day when I was engaging in a friendly
>> competition with a FreeBSD hacker on improving the serial driver for
>> the 8250 chip (with no FIFO's!), and I shared with him my idea of
>> using a pair of ring buffers that would get flipped back and forth
>> between the interrupt handler and the tty "bottom-half" (read:
>> software interrupt) handler, and I was told that clists were handed
>> down from Olympus by the AT&T/Unix Gods and he could never get that
>> kind of change into the FreeBSD tty layer. Of course, I was free to
>> make all of the radical changes to Linux's tty layer --- and I did,
>> all in the name of the number of 115kbaud connections that could be
>> handled on a single 40 MHz 386 processor...
> I remember being pleasantly surprised that Linus/Linux was open to
> that sort of change. I get why the traditional Unix shops resisted
> be wacks like that but they went too far and it prevented good work.
> Linux seemed far more willing to realize that they had it wrong and
> there was a better way. That was refreshing.
> Of course they got beat up for it with "Linux is stable/compatible".
At Fortune Systems Dave Yost was able to achieve full-duplex 9600
baud speed on up to 5 ports in V7 Unix without changing the clist
design. This on a 5.6Mhz machine (with 4 cycle memory). The trick
was to specialize interrupt handlers for each port.
More information about the TUHS