Just looking through J3/18-007r1 ...
11.6.7 says
R1171 event-variable is scalar-variable
C1176 (R1171) An event-variable shall be of type EVENT_TYPE from
the intrinsic module ISO_FORTRAN- ENV (16.10.2).
Clear so far.
11.6.8 tells me
C1177 (R1172) The event-variable in an event-wait-stmt shall not
be coindexed.
and
The event-variable shall not depend on the value of stat-variable
or errmsg-variable.
So... does an event-variable have to be a coarray? It would make
sense, certainly, but I see nothing prescribing it. Or am I missing something?
Just looking through J3/18-007r1 ...
11.6.7 says
R1171 event-variable is scalar-variable
C1176 (R1171) An event-variable shall be of type EVENT_TYPE from
the intrinsic module ISO_FORTRAN- ENV (16.10.2).
Clear so far.
11.6.8 tells me
C1177 (R1172) The event-variable in an event-wait-stmt shall not
be coindexed.
and
The event-variable shall not depend on the value of stat-variable
or errmsg-variable.
So... does an event-variable have to be a coarray? It would make
sense, certainly, but I see nothing prescribing it. Or am I missing something?
So... does an event-variable have to be a coarray? It would make
sense, certainly, but I see nothing prescribing it. Or am I missing something?
Thomas Koenig schrieb am Freitag, 25. Februar 2022 um 08:19:06 UTC+1:
Just looking through J3/18-007r1 ...
11.6.7 says
R1171 event-variable is scalar-variable
C1176 (R1171) An event-variable shall be of type EVENT_TYPE from
the intrinsic module ISO_FORTRAN- ENV (16.10.2).
Clear so far.
11.6.8 tells me
C1177 (R1172) The event-variable in an event-wait-stmt shall not
be coindexed.
and
The event-variable shall not depend on the value of stat-variable
or errmsg-variable.
So... does an event-variable have to be a coarray? It would make
sense, certainly, but I see nothing prescribing it. Or am I missing
something?
The requirement that event variables be coarrays is in 16.10.2.10 (C1604).
On Thursday, February 24, 2022 at 11:19:06 PM UTC-8, Thomas Koenig wrote:
(snip)
So... does an event-variable have to be a coarray? It would make
sense, certainly, but I see nothing prescribing it. Or am I missing
something?
I would have thought EVENT variables would be used with
asynchronous I/O, and then would not be coarrays, but then I haven't
tried using asynchronous I/O in Fortran.
The underlying I/O model for many IBM OSs is asynchronous, based on
the ECB, event control block, which I believe is related to event variables in
some high-level languages.
Programs can test the ECB for completion, and then when appropriate,
WAIT on one (or more). You can, for example, wait for only some of a
group of ECB to complete, such as only one, and then handle the completed
on as appropriate. That is, have many outstanding events, and handle them
as they complete.
But I don't see why that should be coarrays.
Sysop: | Keyop |
---|---|
Location: | Huddersfield, West Yorkshire, UK |
Users: | 157 |
Nodes: | 16 (0 / 16) |
Uptime: | 16:32:09 |
Calls: | 3,193 |
Files: | 10,512 |
Messages: | 2,978,769 |