|
From: Nicholas N. <nj...@cs...> - 2005-11-18 20:24:04
|
Hi, Addrcheck has not worked since Valgrind 2.4.0. A number of people have complained about this, and we'd like to reinstate it, or reinstate something like it. On the other hand, it would be great if we could get rid of Addrcheck and just use Memcheck because that would be less code to maintain. So I'd like to know why people who use Addrcheck do so in preference to Memcheck. I can think of four possible reasons: 1. Addrcheck is faster 2. Addrcheck uses less memory 3. Addrcheck does not issue undefined value error messages 4. Something else I haven't thought of Which of these reasons are important will affect possible solutions. For example, we're considering the idea of using a compressed V bit scheme that would reduce the overhead of shadow memory in Memcheck (the main memory overhead) from 112.5% to about 25%. (Addrcheck's overhead is 12.5%.) If reason (2) is the main reason people prefer Addrcheck, then compressed V bits should be enough of an improvement to Memcheck that Addrcheck is no longer necessary. But if reason (1) is the main reason, then maybe not. So, Addrcheck users, please speak up! There are a lot of you... the recent survey (which I've almost finished analysing) shows that around 25% of Valgrind users use Addrcheck. Tell me which one (or more) of the above reasons is important to you. Your feedback will help with the task of getting Addrcheck or an acceptable equivalent working. If you don't want to reply to the list, please email me privately. Thanks. Nick |
|
From: Brian C. <cr...@fi...> - 2005-11-18 20:32:38
|
For me it was the memory issue, and in 2.4 the speed issue. Memcheck in 3.0 seems significantly faster anyway, so those are all I care about. #3 can be fixed with suppressions, so I say ditch addrcheck. :) -- Brian Nicholas Nethercote wrote: > Hi, > > Addrcheck has not worked since Valgrind 2.4.0. A number of people have > complained about this, and we'd like to reinstate it, or reinstate > something like it. On the other hand, it would be great if we could get > rid of Addrcheck and just use Memcheck because that would be less code > to maintain. > > So I'd like to know why people who use Addrcheck do so in preference to > Memcheck. I can think of four possible reasons: > > 1. Addrcheck is faster > 2. Addrcheck uses less memory > 3. Addrcheck does not issue undefined value error messages > 4. Something else I haven't thought of > > Which of these reasons are important will affect possible solutions. For > example, we're considering the idea of using a compressed V bit scheme > that would reduce the overhead of shadow memory in Memcheck (the main > memory overhead) from 112.5% to about 25%. (Addrcheck's overhead is > 12.5%.) If reason (2) is the main reason people prefer Addrcheck, then > compressed V bits should be enough of an improvement to Memcheck that > Addrcheck is no longer necessary. But if reason (1) is the main reason, > then maybe not. > > So, Addrcheck users, please speak up! There are a lot of you... the > recent survey (which I've almost finished analysing) shows that around > 25% of Valgrind users use Addrcheck. Tell me which one (or more) of the > above reasons is important to you. Your feedback will help with the > task of getting Addrcheck or an acceptable equivalent working. If you > don't want to reply to the list, please email me privately. > > Thanks. > > Nick > > > ------------------------------------------------------- > This SF.Net email is sponsored by the JBoss Inc. Get Certified Today > Register for a JBoss Training Course. Free Certification Exam > for All Training Attendees Through End of 2005. For more info visit: > http://ads.osdn.com/?ad_id=7628&alloc_id=16845&op=click > _______________________________________________ > Valgrind-users mailing list > Val...@li... > https://lists.sourceforge.net/lists/listinfo/valgrind-users > > > |
|
From: <Woe...@on...> - 2005-11-19 09:36:43
|
On Friday 18 November 2005 21:23, Nicholas Nethercote wrote: > Hi, > > So I'd like to know why people who use Addrcheck do so in preference > to Memcheck. I can think of four possible reasons: > > 1. Addrcheck is faster > 2. Addrcheck uses less memory > 3. Addrcheck does not issue undefined value error messages > 4. Something else I haven't thought of I only used Addrcheck when I was looking for memory leaks. Then nothing=20 but 1. was important for me. Cheers, Andr=E9 |
|
From: Dennis L. <pla...@in...> - 2005-11-19 11:47:24
|
At 21:23 18.11.2005, Nicholas Nethercote wrote: >Hi, > >So I'd like to know why people who use Addrcheck do so in preference to >Memcheck. I can think of four possible reasons: > > 1. Addrcheck is faster > 2. Addrcheck uses less memory > 3. Addrcheck does not issue undefined value error messages > 4. Something else I haven't thought of I was always using memcheck for running my programs, and only switched to addrcheck when memcheck failed, mostly due to out of memory with memcheck. So 2. was the most important reason to me to use it. I never felt for my programs running under addrcheck was really much faster, but that was probably due to much i/o wait. For 3. Id say that would no be (the only) reason to resurrect addrcheck, just run memcheck with --suppress-all-underfined-value-error-messages or whatever. Whenever I had to use addrcheck, it felt harder to track down actual errors, as addrcheck output seemed not so good as memchecks. Personally, I would not miss addrcheck much.. greets Dennis Carpe quod tibi datum est |
|
From: Dirk M. <dm...@gm...> - 2005-11-20 14:53:26
|
On Friday 18 November 2005 21:23, Nicholas Nethercote wrote: > 1. Addrcheck is faster Biggest reason. you want to quickly find out where you corrupt the heap. undefined value access check is usually not that important. Dirk |