Midnight Commander is very slow when starting and changing directories
I'm using midnight commander with Cygwin 64 bit on Windows 10 64 bit.
After moving to a new machine and a clean reinstall of windows and
cygwin, my midnight commander started to get very laggy - starting mc,
any change of directory or exiting from finding files (to refresh list
of files in the two panes) takes long seconds. Restarting doesn't
help.
I have no idea what could be the reason for these hiccups. MC normally
works fine: arrow presses work well, viewing and editing is instant,
but this one thing takes up to 10 seconds - changing directories.
Searching in files has a similar lag, but only once (ie not in every
`chdir` it could have done). In non-mc cygwin usage this issue doesn't
happen.
The task manager doesn't show any abnormally high CPU or memory usage
on any process. I have `cygserver` running, using `bash` as my shell
and have no fancy `PS1` settings, my `$CYGWIN` is only
`winsymlinks:native`.
My environment:
* Windows 10 pro 64 bit, up to date
* Not connected to a domain, no network drives
* No anti virus software apart from the one built in Windows
* Two hard drives (system on SSD (C:), D: is a SSHD), quite a few
directory junctions from C: to C:, no loops
* I'm usually using fatty as the terminal, but mintty has the same issues
* GNU Midnight Commander 4.8.17
* cygserver is installed
Here is a strace log, maybe someone has an idea what did go wrong?
http://pastebin.pl/view/9b839815
(recorded with `LOCALE=C strace -q -o /tmp/mc-log mc /tmp`)
In the recorded session I have:
* started mc in `/tmp` directory
* entered `/tmp/NVIDIA Corporation` directory
* went back up (`..`)
* entered `/tmp/mc-chanibal` directory
* went back up (`..`)
* exited with F10
You can see that normalize_posix_path takes a lot of time, but trying
to force the same call with cygpath fails to take so long (more
later).
Additional timing information:
A find of all directories in my `$HOME` takes 4.6 seconds:
$ time find ~ -type d | wc -l
(snipped two directories without permissions)
19919
real0m4.626s
user0m0.796s
sys 0m3.765s
A `ls -al` on all of these 19919 directories took a lot of time, but
the average was fairly low:
time (find ~ -type d | while read f; do { time ls -al $f; } \
|& awk '$1 == "real" { printf $2 }' \
| awk -Fm '{printf( "%2.4f", $1 * 60 + $2 )}'; \
echo -e "\t$f"; done \
) >/tmp/ls-time-log ;
(snipped two dirs without permissions)
real12m21.870s
user7m4.576s
sys 16m34.446s
The average time to do an `ls -al` is 0.020s, min is 0.017, max is 0.154:
$ awk 'BEGIN { sum=0; min=; max=0; minf=""; maxf=""; } \
$1 { sum += $1; \
if(min > $1) { min = $1; minf = $2; }; \
if(max < $1) { max = $1; maxf = $2; } } \
END { print "avg: " sum/NR; print "min: " min " " \
minf; print "max: " max " " maxf }' http://superuser.com/questions/1097601/midnight-commander-sometimes-very-slow
-- KB
--
Problem reports: http://cygwin.com/problems.html
FAQ: http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Re: Midnight Commander is very slow when starting and changing directories
On 08/07/2016 16:13, Krzysztof Bociurko wrote: I'm using midnight commander with Cygwin 64 bit on Windows 10 64 bit. After moving to a new machine and a clean reinstall of windows and cygwin, my midnight commander started to get very laggy - starting mc, any change of directory or exiting from finding files (to refresh list of files in the two panes) takes long seconds. Restarting doesn't help. I have also posted this issue to Super User, but after some discussion with user Matzeri, he suggested for me to go on this list with this problem. SU link: http://superuser.com/questions/1097601/midnight-commander-sometimes-very-slow As mentioned on https://cygwin.com/problems.html "Run cygcheck -s -v -r > cygcheck.out and include that file as an attachment in your report. Please do not compress or otherwise encode the output. Just attach it as a straight text file so that it can be easily viewed. " please attach a copy of cygcheck.out Regards Marco -- Problem reports: http://cygwin.com/problems.html FAQ: http://cygwin.com/faq/ Documentation: http://cygwin.com/docs.html Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Re: Midnight Commander is very slow when starting and changing directories
On 08/07/2016 16:13, Krzysztof Bociurko wrote: I'm using midnight commander with Cygwin 64 bit on Windows 10 64 bit. After moving to a new machine and a clean reinstall of windows and cygwin, my midnight commander started to get very laggy - starting mc, any change of directory or exiting from finding files (to refresh list of files in the two panes) takes long seconds. Restarting doesn't help. I have no idea what could be the reason for these hiccups. MC normally works fine: arrow presses work well, viewing and editing is instant, but this one thing takes up to 10 seconds - changing directories. Searching in files has a similar lag, but only once (ie not in every `chdir` it could have done). In non-mc cygwin usage this issue doesn't happen. The task manager doesn't show any abnormally high CPU or memory usage on any process. I have `cygserver` running, using `bash` as my shell and have no fancy `PS1` settings, my `$CYGWIN` is only `winsymlinks:native`. My environment: * Windows 10 pro 64 bit, up to date * Not connected to a domain, no network drives * No anti virus software apart from the one built in Windows * Two hard drives (system on SSD (C:), D: is a SSHD), quite a few directory junctions from C: to C:, no loops * I'm usually using fatty as the terminal, but mintty has the same issues * GNU Midnight Commander 4.8.17 * cygserver is installed two ideas to try: - update cygwin package as 2.5.2 versus the 2.5.1 that you are running has few bug fixes that could change the responsiveness. (unlike to work but worth to try) - disable the windows anti virus. Experience in the past with Windows Defender were disappointing https://cygwin.com/faq/faq.html#faq.using.bloda Regards Marco -- Problem reports: http://cygwin.com/problems.html FAQ: http://cygwin.com/faq/ Documentation: http://cygwin.com/docs.html Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Re: Midnight Commander is very slow when starting and changing directories
Disabled windows defender real time and cloud based protection, restarted Cygwin. No change. Updating cygwin will have to wait a bit. 2016-07-08 16:49 GMT+02:00 Marco Atzeri : > On 08/07/2016 16:13, Krzysztof Bociurko wrote: >> >> I'm using midnight commander with Cygwin 64 bit on Windows 10 64 bit. >> >> After moving to a new machine and a clean reinstall of windows and >> cygwin, my midnight commander started to get very laggy - starting mc, >> any change of directory or exiting from finding files (to refresh list >> of files in the two panes) takes long seconds. Restarting doesn't >> help. >> >> I have no idea what could be the reason for these hiccups. MC normally >> works fine: arrow presses work well, viewing and editing is instant, >> but this one thing takes up to 10 seconds - changing directories. >> Searching in files has a similar lag, but only once (ie not in every >> `chdir` it could have done). In non-mc cygwin usage this issue doesn't >> happen. >> >> The task manager doesn't show any abnormally high CPU or memory usage >> on any process. I have `cygserver` running, using `bash` as my shell >> and have no fancy `PS1` settings, my `$CYGWIN` is only >> `winsymlinks:native`. >> >> My environment: >> * Windows 10 pro 64 bit, up to date >> * Not connected to a domain, no network drives >> * No anti virus software apart from the one built in Windows >> * Two hard drives (system on SSD (C:), D: is a SSHD), quite a few >> directory junctions from C: to C:, no loops >> * I'm usually using fatty as the terminal, but mintty has the same issues >> * GNU Midnight Commander 4.8.17 >> * cygserver is installed >> > > two ideas to try: > > - update cygwin package as 2.5.2 versus the 2.5.1 that you are running > has few bug fixes that could change the responsiveness. > (unlike to work but worth to try) > > - disable the windows anti virus. > Experience in the past with Windows Defender were disappointing > https://cygwin.com/faq/faq.html#faq.using.bloda > > > Regards > Marco > > > > -- > Problem reports: http://cygwin.com/problems.html > FAQ: http://cygwin.com/faq/ > Documentation: http://cygwin.com/docs.html > Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple > -- Pozdrawiam, Krzysiek Bociurko +48 664 964 967 -- Problem reports: http://cygwin.com/problems.html FAQ: http://cygwin.com/faq/ Documentation: http://cygwin.com/docs.html Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Re: cygstart.exe can't open file:///C:/
Thanks for clarification, guys! Indeed it was my bad for including
customary three forward slashes. Both file://C:/ and file://./C:/ do
work with cygstart.
Appreciate the help.
On 7 July 2016 at 20:43, Andrey Repin wrote:
> Greetings, Brian Inglis!
>
>> Andrey Repin writes:
>>>Brian Inglis writes:
cygstart file://C:/ works - read the MS DN and MS KB articles on file URIs
and shlwapi
>>> Which isn't quite right. "file:" is a protocol, "//" is the foreign host
>> mark, "[.]/" is "current host's filesystem root".
>>> So, I guess, the CORRECT solution (or, rather, workaround) would be an
>> explicit "." in host name.
>>> cygstart "file://./C:/"
>>> Works here. Please try it yourself.
>
>> MS approach makes some sense, as the RFCs e.g. 3986 define what you call the
>> the "host" as the namespace authority. In Unix systems, you have only one
>> unified local namespace (even though the mounted filesystems can have
>> radically different namespace rules e.g. fat, ufs, ext?, and the RFCs state
>> the authority may be delegated, so the rules can change along the path),
>> whereas on Windows, each device represents (possibly virtual e.g. subst)
>> separate filesystem namespaces.
>
>> Where MS approach makes no sense, is that . is a (MS) kludge which works,
>> but other local synonyms: null/nothing, localhost, 127.0.0.1, [::1] do not,
>> whereas $BROWSER file://{,.,localhost,127.0.0.1,::1}/C{:,\|} displays
>> identical contents, differing only in whether a : or | follows the drive
>> letter in the address for each tab.
>
> file://localhost/C:/ works, at least for CMD call. Not for cygstart, though.
> Using IP, of course, does not, which, again, makes sense.
> Browsers, on the other hand, often have their own protocol translation, so you
> can't quite compare their behavior to native API calls.
>
>> I dealt with a Windows product where file: (but not ftp, http, or https) had
>> to have an initial cap File: to work. The vendor accepted a bug report but
>> made it a doc issue rather than doing a non-compliance fix. The company
>> and/or products were traded annually like an end of career player!
>
> *sigh*
>
>
> --
> With best regards,
> Andrey Repin
> Thursday, July 7, 2016 20:38:24
>
> Sorry for my terrible english...
>
>
> --
> Problem reports: http://cygwin.com/problems.html
> FAQ: http://cygwin.com/faq/
> Documentation: http://cygwin.com/docs.html
> Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
>
--
Problem reports: http://cygwin.com/problems.html
FAQ: http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Re: scanf and m modifier
On 7/8/2016 1:16 PM, Owen Leibman wrote: The m modifier is handled just fine by the compiler, but the results of using it are different on Cygwin than on several other systems I've tried. Here is a sample program: Dear Owen -- The first question I'd ask is: What versions of gcc and C libraries are we talking about? IIRC, the m feature is relatively new (relative being, umm, relative), so which gcc you're using, and the libraries, too, may matter. Regards -- Eliot Moss -- Problem reports: http://cygwin.com/problems.html FAQ: http://cygwin.com/faq/ Documentation: http://cygwin.com/docs.html Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
[ANNOUNCEMENT] fish 2.3.1-1
fish 2.3.1-1 is now available in Cygwin. This is an upstream release with bug fixes and minor new features. See http://fishshell.com/release_notes.html for the list. fish 2.2.0-4 is still the previous release. This is in case anyone isn't ready to upgrade to version 2.3 yet, but I'll probably let 2.2 age out of the distribution at the next update. fish is the friendly interactive shell. It's a Unix shell that focuses on interactive use, discoverability, and user friendliness. The design goal of fish is to give the user a rich set of powerful features in a way that is easy to discover, remember, and use. Home page: http://fishshell.com Andrew E. Schulman *** To update your installation, click on the "Install Cygwin now" link on the http://cygwin.com/ web page. This downloads setup.exe to your system. Then, run setup and answer all of the questions. *** CYGWIN-ANNOUNCE UNSUBSCRIBE INFO *** If you want to unsubscribe from the cygwin-announce mailing list, look at the "List-Unsubscribe: " tag in the email header of this message. Send email to the address specified there. It will be in the format: cygwin-announce-unsubscribe-you=yourdomain.com_at_cygwin.com If you need more information on unsubscribing, start reading here: http://cygwin.com/lists.html#subscribe-unsubscribe Please read *all* of the information on unsubscribing that is available starting at this URL. -- Problem reports: http://cygwin.com/problems.html FAQ: http://cygwin.com/faq/ Documentation: http://cygwin.com/docs.html Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Re: scanf and m modifier
On 07/08/2016 11:22 AM, Eliot Moss wrote:
> On 7/8/2016 1:16 PM, Owen Leibman wrote:
>> The m modifier is handled just fine by the compiler, but
>> the results of using it are different on Cygwin than on several other
>> systems
>> I've tried. Here is a sample program:
>
> Dear Owen -- The first question I'd ask is: What versions of gcc
> and C libraries are we talking about? IIRC, the m feature is
> relatively new (relative being, umm, relative), so which gcc you're
> using, and the libraries, too, may matter.
scanf("%ms") is a POSIX extension, not required by C99. I suspect the
answer is that no one has implemented it in newlib yet; patches welcome.
--
Eric Blake eblake redhat com+1-919-301-3266
Libvirt virtualization library http://libvirt.org
signature.asc
Description: OpenPGP digital signature
[ANNOUNCEMENT] libwebp 0.4.4-1
The following packages have been uploaded to the Cygwin distribution: * libwebp-0.4.4-1 * libwebp5-0.4.4-1 * libwebpmux1-0.4.4-1 * libwebpdemux1-0.4.4-1 * libwebpdecoder1-0.4.4-1 * libwebp-devel-0.4.4-1 * mingw64-i686-libwebp-0.4.4-1 * mingw64-x86_64-libwebp-0.4.4-1 WebP is a new image format that provides lossy compression for photographic images. A WebP file consists of VP8 image data, and a container based on RIFF. The standalone libwebp library serves as a reference implementation for the WebP specification. Webmasters and web developers can use the WebP image format to create smaller and better looking images that can help make the web faster. This is the latest update for the 0.4 branch. -- Yaakov -- Problem reports: http://cygwin.com/problems.html FAQ: http://cygwin.com/faq/ Documentation: http://cygwin.com/docs.html Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
[ANNOUNCEMENT] libvpx 1.5.0-1
The following packages have been uploaded to the Cygwin distribution: * libvpx-1.5.0-1 * libvpx3-1.5.0-1 * libvpx-devel-1.5.0-1 * libvpx-doc-1.5.0-1 * libvpx-utils-1.5.0-1 * mingw64-i686-libvpx-1.5.0-1 * mingw64-x86_64-libvpx-1.5.0-1 WebM is an open, royalty-free, media file format designed for the web. WebM defines the file container structure, video and audio formats. WebM files consist of video streams compressed with the VP8 video codec and audio streams compressed with the Vorbis audio codec. The WebM file structure is based on the Matroska container. This is an update to the latest upstream release, which includes an ABI version bump. -- Yaakov -- Problem reports: http://cygwin.com/problems.html FAQ: http://cygwin.com/faq/ Documentation: http://cygwin.com/docs.html Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
REQUEST: Add php-pcntl to setup
Cygwin can handle signals so it would be useful for PHP to handle signals as well. http://php.net/manual/en/pcntl.installation.php signature.asc Description: OpenPGP digital signature
