The default of 128kB is unchanged, but the upper limit is changed from
32 blocks to 128 blocks, unless the operating system's IOV_MAX is too
low. Some other RDBMSes seem to cap their multi-block buffer pool I/O
around this number, and it seems useful to allow experimentation.
The concrete change is to our definition of PG_IOV_MAX, which provides
the maximum for io_combine_limit and io_max_combine_limit. It also
affects a couple of other places that work with arrays of struct iovec
or smaller objects on the stack, so we still don't want to use the
system IOV_MAX directly without a clamp: it is not under our control and
likely to be 1024. 128 seems acceptable for our current usage.
For Windows, we can't use real scatter/gather yet, so we continue to
define our own IOV_MAX value of 16 and emulate preadv()/pwritev() with
loops. Someone would need to research the trade-offs of raising that
number.
NB if trying to see this working: you might temporarily need to hack
BAS_BULKREAD to be bigger, since otherwise the obvious way of "a very
big SELECT" is limited by that for now.
Suggested-by: Tomas Vondra <tomas@vondra.me>
Discussion: https://postgr.es/m/CA%2BhUKG%2B2T9p-%2BzM6Eeou-RAJjTML6eit1qn26f9twznX59qtCA%40mail.gmail.com
<!-- doc/src/sgml/README.non-ASCII -->
Representation of non-ASCII characters
--------------------------------------
Find non-ASCII characters using:
grep --recursive --color='auto' -P '[\x80-\xFF]' .
Convert to HTML4 named entity (&) escapes
-----------------------------------------
We support several output formats:
* html (supports all Unicode characters)
* man (supports all Unicode characters)
* pdf (supports only Latin-1 characters)
* info
While some output formatting tools support all Unicode characters,
others only support Latin-1 characters. Specifically, the PDF rendering
engine can only display Latin-1 characters; non-Latin-1 Unicode
characters are displayed as "###".
Therefore, in the SGML files, we only use Latin-1 characters. We
typically encode these characters as HTML entities, e.g., Álvaro.
It is also possible to safely represent Latin-1 characters in UTF8
encoding for all output formats.
Do not use UTF numeric character escapes (&#nnn;).
HTML entities
official: http://www.w3.org/TR/html4/sgml/entities.html
one page: http://www.zipcon.net/~swhite/docs/computers/browsers/entities_page.html
other lists: http://www.zipcon.net/~swhite/docs/computers/browsers/entities.html
http://www.zipcon.net/~swhite/docs/computers/browsers/entities_page.html
https://en.wikipedia.org/wiki/List_of_XML_and_HTML_character_entity_references
06fb5612c9