Re: Architectural implications of locate mode I/O

Liste des GroupesRevenir à c arch 
Sujet : Re: Architectural implications of locate mode I/O
De : terje.mathisen (at) *nospam* tmsw.no (Terje Mathisen)
Groupes : comp.arch
Date : 05. Jul 2024, 09:59:39
Autres entêtes
Organisation : A noiseless patient Spider
Message-ID : <v6895b$36v5v$1@dont-email.me>
References : 1 2 3 4 5 6
User-Agent : Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:91.0) Gecko/20100101 Firefox/91.0 SeaMonkey/2.53.18.2
Lawrence D'Oliveiro wrote:
On Tue, 02 Jul 2024 17:36:50 -1000, Lynn Wheeler wrote:
 
When doing IBM's HA/CMP and working with major RDBMS vendors on cluster
scaleup in late 80s/early 90s, there was lots of references to POSIX
light-weight threads ...
 Threads were all the rage in the 1990s. People were using them for
everything. One language (Java) absorbed threading right into its core DNA
(where is the locking API? Oh, it’s attached to the base “Object” type
itself!).
 People backed off a bit after that. Nowadays we see a revival of the
“coroutine” idea, where preemption only happens at explicit “await”
points. For non-CPU-intensive workloads, this is much easier to cope with.
 
... and asynchronous I/O for RDBMS (with no buffer > copies) and the
RDBMS managing large record cache.
 This is why POSIX has the disk-oriented “aio” API, for the diehard DBMS
folks. Linux also added “io_uring”, for high performance but not disk-
specific I/O.
Really old PC printers (dot matrix or similar) still had a line buffer worth of on-device memory, enough so that the CPU could sit in a buzy loop sending bytes over the Centronix interface until it got the "I'm full" status bit back, or N bytes had been sent. At that point my disk spooler code would back off and let the next timer interrupt check to see if there was both more print data to be sent and the printer signalled that it was ready to receive more data.
The original serial ports had no buffer at all, so you had to use interrupts for both sending and receiving if you wanted to do anything else while communicating. Around 1984 the serial ports gained a 16-byte buffer, so at that time it made sense to use a hybrid approach, filling/emptying the buffer on each interrupt, while reducing the interrupt load by an order of magnitude.
I.e. acting somewhat like a channel program, but using the main/only cpu to do all the work in the background.
Terje
--
- <Terje.Mathisen at tmsw.no>
"almost all programming can be viewed as an exercise in caching"

Date Sujet#  Auteur
2 Jul 24 * Architectural implications of locate mode I/O64John Levine
2 Jul 24 +* Re: Architectural implications of locate mode I/O38Thomas Koenig
3 Jul 24 i+- Re: Architectural implications of locate mode I/O1MitchAlsup1
3 Jul 24 i+* Re: Architectural implications of locate mode I/O17MitchAlsup1
3 Jul 24 ii+* Re: Architectural implications of locate mode I/O6Lynn Wheeler
3 Jul 24 iii+* Re: Architectural implications of locate mode I/O2Lawrence D'Oliveiro
5 Jul 24 iiii`- Re: Architectural implications of locate mode I/O1Terje Mathisen
3 Jul 24 iii+* Re: Architectural implications of locate mode I/O2MitchAlsup1
3 Jul 24 iiii`- Re: Architectural implications of locate mode I/O1Michael S
3 Jul 24 iii`- Re: Architectural implications of locate mode I/O1Lynn Wheeler
3 Jul 24 ii+* Re: Architectural implications of locate mode I/O9Bill Findlay
3 Jul 24 iii`* Re: Architectural implications of locate mode I/O8Michael S
3 Jul 24 iii +- Re: Architectural implications of locate mode I/O1Bill Findlay
3 Jul 24 iii `* Re: Architectural implications of locate mode I/O6Stephen Fuld
3 Jul 24 iii  `* Re: Architectural implications of locate mode I/O5MitchAlsup1
4 Jul 24 iii   `* Re: wisdom of the ancients, was Architectural implications of locate mode I/O4John Levine
4 Jul 24 iii    +* Re: wisdom of the ancients, was Architectural implications of locate mode I/O2John Savard
4 Jul 24 iii    i`- Re: wisdom of the ancients, was Architectural implications of locate mode I/O1John Savard
4 Jul 24 iii    `- Re: wisdom of the ancients, was Architectural implications of locate mode I/O1MitchAlsup1
3 Jul 24 ii`- Re: Architectural implications of locate mode I/O1Stephen Fuld
3 Jul 24 i`* Re: Architectural implications of locate mode I/O and channels19John Levine
3 Jul 24 i +* Re: Architectural implications of locate mode I/O and channels9Anton Ertl
3 Jul 24 i i+* Re: Architectural implications of locate mode I/O and channels6Michael S
3 Jul 24 i ii+* Re: Architectural implications of locate mode I/O and channels4MitchAlsup1
4 Jul 24 i iii`* Re: Architectural implications of locate mode I/O and channels3Michael S
4 Jul 24 i iii +- Re: Architectural implications of locate mode I/O and channels1MitchAlsup1
5 Jul 24 i iii `- Re: Architectural implications of locate mode I/O and channels1Joe Pfeiffer
3 Jul 24 i ii`- Re: Architectural implications of locate mode I/O and channels1Michael S
3 Jul 24 i i+- Re: Architectural implications of locate mode I/O and channels1Robert Swindells
3 Jul 24 i i`- Re: Architectural implications of locate mode I/O and channels1John Levine
4 Jul 24 i `* Re: Architectural implications of locate mode I/O and channels9Thomas Koenig
4 Jul 24 i  `* Re: Architectural implications of locate mode I/O and channels8MitchAlsup1
4 Jul 24 i   `* Re: Architectural implications of locate mode I/O and channels7Thomas Koenig
5 Jul 24 i    +* Re: Architectural implications of locate mode I/O and channels5John Levine
5 Jul 24 i    i`* Re: Architectural implications of locate mode I/O and channels4MitchAlsup1
7 Jul 24 i    i `* Re: Architectural implications of locate mode I/O and channels3Paul A. Clayton
7 Jul 24 i    i  +- Re: Architectural implications of locate mode I/O and channels1Lynn Wheeler
7 Jul 24 i    i  `- Re: patents, Architectural implications of locate mode I/O and channels1John Levine
5 Jul 24 i    `- Re: Architectural implications of locate mode I/O and channels1MitchAlsup1
3 Jul 24 +- Re: Architectural implications of people who don’t understand locate mode I/O1Lawrence D'Oliveiro
5 Jul 24 `* Re: Architectural implications of locate mode I/O24Joe Pfeiffer
5 Jul 24  +* Re: Architectural implications of locate mode I/O21John Levine
6 Jul 24  i+* Re: Architectural implications of locate mode I/O18Lynn Wheeler
6 Jul 24  ii+- Re: Architectural implications of locate mode I/O1MitchAlsup1
6 Jul 24  ii+* Re: Architectural implications of locate mode I/O14Stephen Fuld
6 Jul 24  iii+* Re: ancient disks, Architectural implications of locate mode I/O7John Levine
6 Jul 24  iiii`* Re: ancient disks, Architectural implications of locate mode I/O6Stephen Fuld
6 Jul 24  iiii `* Re: ancient disks, Architectural implications of locate mode I/O5John Levine
7 Jul 24  iiii  `* Re: ancient disks, Architectural implications of locate mode I/O4Stephen Fuld
7 Jul 24  iiii   `* Re: ancient disks, Architectural implications of locate mode I/O3Thomas Koenig
7 Jul 24  iiii    `* Re: ancient disks, Architectural implications of locate mode I/O2Thomas Koenig
7 Jul 24  iiii     `- Re: ancient disks, Architectural implications of locate mode I/O1Stephen Fuld
6 Jul 24  iii+- Re: Architectural implications of locate mode I/O1Lynn Wheeler
1 Aug 24  iii`* Re: Architectural implications of locate mode I/O5Lawrence D'Oliveiro
1 Aug 24  iii `* Re: Architectural implications of locate mode I/O4Stephen Fuld
1 Aug 24  iii  +* Re: Architectural implications of locate mode I/O2John Levine
2 Aug 24  iii  i`- Re: Architectural implications of locate mode I/O1Stephen Fuld
11 Aug 24  iii  `- Re: Architectural implications of locate mode I/O1Lawrence D'Oliveiro
8 Jul 24  ii`* Re: Architectural implications of locate mode I/O2Lawrence D'Oliveiro
8 Jul 24  ii `- Re: Architectural implications of locate mode I/O1Lynn Wheeler
6 Jul 24  i`* Re: Architectural implications of locate mode I/O2MitchAlsup1
6 Jul 24  i `- Re: Old printers, Architectural implications of locate mode I/O1John Levine
8 Jul 24  `* Re: Architectural implications of locate mode I/O2Lawrence D'Oliveiro
1 Aug 24   `- Re: Architectural implications of locate mode I/O1MitchAlsup1

Haut de la page

Les messages affichés proviennent d'usenet.

NewsPortal