Re: Slow stat (was Startup speed: Still too many stats?)
- First Post
- Replies
- Stats
-
Go to
- ----- 2024 -----
- April
- March
- February
- January
- ----- 2023 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2022 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2021 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2020 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2019 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2018 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2017 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2016 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2015 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2014 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2013 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2012 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2011 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2010 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2009 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2008 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2007 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2006 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2005 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2004 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2003 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2002 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2001 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2000 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 1999 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 1998 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
Darryl Okahata <darrylo(a)sr.hp.com> writes:
Hrvoje Niksic <hniksic(a)srce.hr> wrote:
> > Note, however, that dired in XEmacs is pretty inefficient. With
> > the current dired, most of the time is spent doing an
"indent-region"
> > (and not obtaining the actual directory contents).
>
> No, it's not. We've already been over this. Please rebuild your
> XEmacs without error checking and you should get a massive speed
> increase in dired.
I haven't tried this, but I guess I have to. ;-)
Please, please do. For me, the error-checkingless rebuild blasted
`indent-region' totally out of the profile trace.
At the time, I spent an hour tracing the `indent-region' slowness by
various unholy means (including but not limited to rewriting several
`indent.el' functions to C), and finally traced it to the code that
simply inserts the characters into the buffer. The reason why this
code was slow was that, when building with ERROR_CHECK_EXTENTS,
adjusting the extents appears to cause sledgehammer extent checks,
which grind things to a halt.
For this reason, the profiling code now prints a message that warns
you of likely nonsensical result with error-checking builds. The
worst thing is: not only are the results nonsensical, but they get you
on the totally wrong track!
It's only fortunate that you listened to your intuition rather than
the profiler output, and wrote the dired-in-C simulation for Windows.
> > To speed things up, try applying the appended patch to
dired.el (I
> > just submitted this to the dired folks).
>
> That patch created problems in the past.
I think you're confusing my patch with someone else's.
Quite possible, sorry.
--
Hrvoje Niksic <hniksic(a)srce.hr> | Student at FER Zagreb, Croatia
--------------------------------+--------------------------------
I luv the smell of nature in the morning. Smells like manure!
Darryl Okahata
Friday, 5 June
Fri, 5 Jun
8:08 a.m.
New subject: Slow stat (was Startup speed: Still too many stats?)
Hrvoje Niksic <hniksic(a)srce.hr> wrote:
> > No, it's not. We've already been over this.
Please rebuild your
> > XEmacs without error checking and you should get a massive speed
> > increase in dired.
>
> I haven't tried this, but I guess I have to. ;-)
Please, please do. For me, the error-checkingless rebuild blasted
`indent-region' totally out of the profile trace.
OK, dired's definitely much faster with error checking turned
off. ;-) However, I still hope the dired patch will be accepted by the
dired maintainers, because it would allow fast direds even with error
checking turned on. Here's a sample result from a 200MHz PPro, NT
4.0sp3, dired of "c:/winnt/system32":
-------------------------------------------------------------------------------
Function Name Ticks %/Total Call Count
==================================== ===== ======= ==========
(in garbage collection) 534 19.940
re-search-forward 508 18.969 3718
set-extent-property 248 9.261 32241
mswindows-insert-directory 212 7.916 2
make-extent 187 6.983 9770
dired-activate-permissions 169 6.311 8793
dired-make-permissions-interactive 109 4.070 978
looking-at 107 3.996 6604
dired-insert-set-properties 82 3.062 1
check-menu-syntax 73 2.726 280
indent-to 72 2.689 978
dired-set-text-properties 68 2.539 977
set-extent-face 66 2.465 9770
dired-manual-move-to-end-of-filename 64 2.390 977
indent-rigidly 52 1.942 1
dired-manual-move-to-filename 48 1.792 978
current-indentation 20 0.747 978
dired-insert-headerline 18 0.672 1
string-match 13 0.485 285
vectorp 8 0.299 1912
keywordp 2 0.075 1089
normalize-menu-item-name 2 0.075 179
copy-sequence 2 0.075 12
file-attributes 2 0.075 2
redisplay-echo-area 2 0.075 2
search-forward 2 0.075 1
selected-frame 2 0.075 5
nsubstitute 2 0.075 2
map-extents 2 0.075 5
#<compiled-function (extent maparg) "...(8)" [extent-property extent
dired-file-name] 3> 2 0.075 20
---------------------------------------------------------------------
Total 2678 100.00
One tick = 1 ms
-------------------------------------------------------------------------------
Hmmm. I'll have to check to find out why `mswindows-insert-directory'
is being called twice.
-- Darryl Okahata
Internet: darrylo(a)sr.hp.com
DISCLAIMER: this message is the author's personal opinion and does not
constitute the support, opinion, or policy of Hewlett-Packard, or of the
little green men that have been following him all day.
9451
days inactive
9454
days old
1 comments
2 participants
tags (0)
participants (2)
-
Darryl Okahata -
Hrvoje Niksic