similar to: Ineffective code after loop unrolling with -O3, ok with -Os

Hello, I've been playing around with the really simple example of not cache friendly loop like this: #define N 100 void foo(int** __restrict__ a, int** __restrict__ b) { for (int i = 0; i < N; ++i) for (int j = 0; j < N; ++j) a[j][i] += b[j][i]; } link to compiler explorer:

Hello, Using some Windows applications with WINE on Puppy/Toutou LINUX 4.31, I noticed that links created in the pages of my Outliners (TreeDBNotes, KeyNotes) are ineffective. Idem for links to websites in the help menus of any Windows applications. What's going wrong and what could be done to solve this problem? Thank you! Regards.

Your observation would be right if you are using --checksum which you shouldn't be. Otherwise, unless you are using --whole-file rsync will use its differential algorithm to compare the files. If you are using --progress you will see it step through the file at a faster speed than the network is capable of. If there is no difference then rsync will go very fast to 100% then do any metadata

Do not use --checksum. It has an extremely limited use case. Normally it is much slower than simply re-copying everything. --checksum means checksum every file on both ends (even files that only exist on one end) before doing anything else even if doing so causes a timeout failure. --checksum is the only part of rsync stupid enough to leave one end idle potentially for hours. On 03/14/2018

https://bugzilla.samba.org/show_bug.cgi?id=10163 Summary: rsync -X is ineffective when setting owner/group Product: rsync Version: 3.1.0 Platform: All OS/Version: All Status: NEW Severity: normal Priority: P5 Component: core AssignedTo: wayned at samba.org ReportedBy: mluscon at redhat.com

This is for Dovecot 1.0.rc15 from Debian Backports on 3.1 Stable, x86; Filesystems are local reiserfs; IMAPs client is KMail 1.9.5 Non-default settings in dovecot.conf: protocols = imaps ssl_cert_file = /etc/dovecot/mail.cert.pem ssl_key_file = /etc/dovecot/mail.key.pem mail_location = maildir:/var/mail/ambrosia.plus.com/%n mail_extra_groups = mail first_valid_uid = 1000

Kevin Korb via rsync <rsync at lists.samba.org> (Mi 14 Mär 2018 14:52:55 CET): > Your observation would be right if you are using --checksum which you > shouldn't be. Otherwise, unless you are using --whole-file rsync will > use its differential algorithm to compare the files. If you are using > --progress you will see it step through the file at a faster speed than Ok,

Hi, how does rsync work if it compares two very huge files on two distinct hosts (rsync uses a networked connection, via SSH)? Some observation seems to indicate, that rsync first reads (and checksums?) the remote (destination) side, then, if finished, it reads (and checksums?) the local (source) side and then starts transferring the delta if done with the comparison of these two files on both

Running rsync 2.6.9-1.el4.rf on CentOS 4.4 client and remote server. Backing up user data from 2 different clients using following: su - $HOSTID -c 'rsync -azr --timeout=600 --log-file=$DEBUGFILE --log-file-format="%o %f %b %l %i" --stats --delete --bwlimit=$BANDWDT --rsh="ssh -P ____" $STAGE $TARGET:$TARGETDIR' Using "bytes sent"/"literal data"

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi, I'm very happy with Dovecot and wish the best for its continued development. It's a true testament to brilliant componentised design and Unix philosophy. Configuration made easy because of the way the file is parsed and a single point of startup make it ridiculously straightforward (and lovely) to administer. But of course, we all

Hi, I set a preference order of the codecs to my sip.conf [general] port = 5060 ; Port to bind to bindaddr = 0.0.0.0 ; Address to bind to context = default ; Default for incoming calls of not registered phones disallow = all allow = g729 allow = g723 allow = alaw allow = ulaw Connected a 'Sipura SPA' sip phone to asterisk with g729 as its preferred codec. Problem: asterisk cannot make

According to the smb.conf man page, the "dos filemode" parameter will allow a user who is not the owner of a file to affect permissions changes on a file or folder provided he has write access to that object. This does not appear to be the case. Users with write access via user or group ACEs receive an "Access denied" error when attempting to make ACL changes via windows,

I am trying to control the behavior of the SANN method in optim (R 2.14.1) via control$temp. In my toy tests it works; in my real use, it doesn't. As far as I can tell my code with different temp values is loaded; I even traced into the function that calls optim and verified temp had the value I had set. Could the fact that I have NaN's coming back from the objective function be a

Interesting. By ld do you mean GNU ld? I know GNU ld does "work" with LLVM's gold plugin, but it's an untested combination and not recommended. I wouldn't be surprised if there were some issues around it not passing necessary info to the gold plugin. Teresa On Mon, Sep 16, 2019 at 8:41 AM Michael Woerister <mwoerister at mozilla.com> wrote: > So one interesting

Can you clarify if performance difference is caused by using different linkers at instrumentation build? If that is the case, try dump the sections of the resulting binary and compare __llvm_prf_** sections. Also check the arguments passed to the linker. It should have -u__llvm_profile_runtime to force the profile runtime to be linked in. David On Mon, Sep 16, 2019 at 8:42 AM Michael

Hi, I have some IR module from random generation (mostly ineffective instructions). It has a function with void return, and two function arguments where one is a reference. Therefore, I expect every instruction not altering the value at the 2nd arguments address should be ineffective. Here is the function definition (see below for full ll): define void @_Z27entityMainDataInputCallbackdRd(double

Interestingly, a C version of the same test program [1] compiled with Clang 8 does not have any problems with GNU ld: The `__llvm_prf_data` section is the same size for all three linkers. It must be something specific to the Rust compiler that's going wrong here. [1] https://github.com/michaelwoerister/rust-pgo-test-programs/tree/master/cpp_branch_weights On Tue, Sep 17, 2019 at 3:26 PM

Hi everyone, As part of my work for Mozilla's Low Level Tools team I've implemented PGO in the Rust compiler. The feature is available since Rust 1.37 [1]. However, so far we have not seen any actual performance gains from enabling PGO for Rust code. Performance even seems to drop 1-3% with PGO enabled. I wonder why that is and I'm hoping that someone here might have experience

On Thu, Sep 12, 2019 at 8:18 AM Teresa Johnson <tejohnson at google.com> wrote: > I just have a couple suggestions off the top of my head: > - have you tried using the new pass manager > (-fexperimental-new-pass-manager)? That has access to additional analysis > info during inlining and is able to make more precise PGO based inline > decisions. > (although note the above

Hi! On 7/10/2020 07:17, Shuai Wang wrote: > Hello! > > Thank you very much! Yes, that makes a lot of sense to me. However, just > want to point out two things that are still unclear: > > 1. The output contains a alias set of only one element, for instance: > "must alias, Mod       Pointers: (i32* %y, LocationSize::precise(4))" > > This one really confused