November 2014 - kernelnewbies@kernelnewbies.org

Unaligned memory access in 2.6.39
by rohit Kulkarni Nov. 14, 2014

Nov. 14, 2014

Hi guys, This is my first post on KernelNewbies.I have a CGI binary written in c++ which uses some shared objects which use STL. The code is legacy code, so it contains a lot of #pragma packs. I used to run this code on kernel 2.6.33. and I updated the kernel to 2.6.39 and the same CGI binary is giving segafault. (which works on older kernel) But if I remove #pragma packs it works as expected on later kernel too. I am using intel atom processor. Can anyone give me some clue about what can I do ? I know that removing pragmas is obvious solution but its affecting some other code. thanks,Rohit

1 0

Re: Re Start Fresh
by Adrian Remonda Nov. 14, 2014

Nov. 14, 2014

> to, when submitting a patch, CC the *maintainers* of the subsystem in > question. which you did not do. strike three. > > SIDE NOTE: when submitting trivial patches, one is also encouraged > to send such patches to the email address "trivial(a)kernel.org". i'm > assuming you've never heard of that. le *sigh* ... > > continuing, yes, there is value in correcting even mistakes in > comments, but typically only if there is the possibility of the > comment typo perhaps confusing or misleading someone. here, however, > it's highly unlikely that anyone will read the word "FIMXE" and think, > "gosh, i wonder what that means." but here's the worst part. Why do you keep answer him! He is just playing around with us. This mail list is getting annoying because of this spam. Just ignore him…. Adrian

1 0

Sched_clock() usage
by Vignesh Radhakrishnan Nov. 14, 2014

Nov. 14, 2014

Hi, Is there any place where sched_clock() usage might not be safe. Like lets say early in the boot during driver probe ? Or any things that i need to ensure before calling sched_clokc() ? Thanks and regards, Vignesh Radhakrishnan -- http://vigneshradhakrishnan.blogspot.com/

2 1

Re Start Fresh
by Nick Krause Nov. 13, 2014

Nov. 13, 2014

I am willing to start out completely fresh and willing to learn how to do this correctly now and hopefully improve my rep ... slowly. This is not another email for me trying to state again I am willing to change, I really would like to help me out and am not sure where to start in order to improve my rep. Cheers Nick

4 4

dmatest module issues
by Ulka Vaze Nov. 13, 2014

Nov. 13, 2014

Hi, I am currently testing crystal beach dma on x86-64 on linux. I have kernel version 2.6.32 on it. As per kernel documentation when dmatest module is inserted in /sys/module/dmatest/parameters/ files are created . However in this case files run and timeout are missing. So I am unable to start dmatest eventhough module is loaded Anyone has idea about it? Please help. Regards, -Ulka -- Director L2IT -Learn & Leverage Information Technology Leading Professional Training Institute in Pune www.l2it.com

1 0

How to disable fault-around by debugfs?
by 秦弋戈 Nov. 12, 2014

Nov. 12, 2014

Hi, I'm just the guy who asked a question about the page fault handler days ago. Thanks for telling me about the fault-around feature. You also told me that this feature could be disabled in debugfs, but I still don't know how to achieve it by debugfs. I only know that debugfs is something that developers could put information about kernel there. So could you tell me how to disable fault-around feature by debugfs? Or is it OK to directly modify kernel? I find where it is located in source: 3518 if (vma->vm_ops->map_pages) { 3519 pte = pte_offset_map_lock(mm, pmd, address, &ptl); 3520 do_fault_around(vma, address, pte, pgoff, flags); 3521 if (!pte_same(*pte, orig_pte)) 3522 goto unlock_out; 3523 pte_unmap_unlock(pte, ptl); 3524 } Thank you very much! Best regards

3 3

RE: Content of Kernelnewbies...
by pushpendra.singh＠smartplayin.com Nov. 12, 2014

Nov. 12, 2014

-----Original Message----- From: kernelnewbies-request(a)kernelnewbies.org Sent: Tuesday, 11 November, 2014 10:30pm To: kernelnewbies(a)kernelnewbies.org Subject: Kernelnewbies Digest, Vol 48, Issue 13 Send Kernelnewbies mailing list submissions to kernelnewbies(a)kernelnewbies.org To subscribe or unsubscribe via the World Wide Web, visit http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies or, via email, send a message with subject or body 'help' to kernelnewbies-request(a)kernelnewbies.org You can reach the person managing the list at kernelnewbies-owner(a)kernelnewbies.org When replying, please edit your Subject line so it is more specific than "Re: Contents of Kernelnewbies digest..." Today's Topics: 1. Confusion about __do_fault and MM_ANONPAGES (Joel Fernandes) ---------------------------------------------------------------------- Message: 1 Date: Mon, 10 Nov 2014 16:23:57 -0800 From: Joel Fernandes <agnel.joel(a)gmail.com> Subject: Confusion about __do_fault and MM_ANONPAGES To: Linux Kernel Mailing List <linux-kernel(a)vger.kernel.org>, "kernelnewbies(a)kernelnewbies.org" <kernelnewbies(a)kernelnewbies.org> Cc: Andrew Morton <akpm(a)linux-foundation.org> Message-ID: <CAD%3DGYpYqbTrJ-wTZy%3DUBX1Z5%3DzMGVoaYU8jSa81rrcC4G0271w(a)mail.gmail.com> Content-Type: text/plain; charset=ISO-8859-1 Hi, I have some confusion about when is anon=1 when __do_fault is called. __do_fault seems to be called only when the page fault is of a linear or a non-linear mapping. My understanding is neither of these can be anonymous, so I am confused about the below code fragment: In which case is anon=1 when memory.c::__do_fault is called? In __do_fault, you'll see something like... if (anon) { inc_mm_counter_fast(mm, MM_ANONPAGES); page_add_new_anon_rmap(page, vma, address); } else { Thanks! Joel ------------------------------ _______________________________________________ Kernelnewbies mailing list Kernelnewbies(a)kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies End of Kernelnewbies Digest, Vol 48, Issue 13 *********************************************

1 0

Learning From Mistakes
by nick Nov. 12, 2014

Nov. 12, 2014

Greetings other developers, After the past few months , I have learned how shitty my patches were initially to the LKML. I am willing to start submitting good patches, if people are willing to give me a second try and this is not one of my many emails for trying and getting it wrong. I really want to start building my trust and relationships again with the maintainers and other developers who's time I so ungratefully wasted. Sorry and I appreciate any replies, Nick

1 0

Confusion about __do_fault and MM_ANONPAGES
by Joel Fernandes Nov. 11, 2014

Nov. 11, 2014

Hi, I have some confusion about when is anon=1 when __do_fault is called. __do_fault seems to be called only when the page fault is of a linear or a non-linear mapping. My understanding is neither of these can be anonymous, so I am confused about the below code fragment: In which case is anon=1 when memory.c::__do_fault is called? In __do_fault, you'll see something like... if (anon) { inc_mm_counter_fast(mm, MM_ANONPAGES); page_add_new_anon_rmap(page, vma, address); } else { Thanks! Joel

1 0

Re: Kernelnewbies Digest, Vol 48, Issue 10
by mohanty bhagaban Nov. 10, 2014

Nov. 10, 2014

Hi , How to get access to min kenel git as i need to add some patches for reviews. Bhagaban On Sat, Nov 8, 2014 at 10:30 PM, <kernelnewbies-request(a)kernelnewbies.org> wrote: > Send Kernelnewbies mailing list submissions to > kernelnewbies(a)kernelnewbies.org > > To subscribe or unsubscribe via the World Wide Web, visit > http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies > or, via email, send a message with subject or body 'help' to > kernelnewbies-request(a)kernelnewbies.org > > You can reach the person managing the list at > kernelnewbies-owner(a)kernelnewbies.org > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of Kernelnewbies digest..." > > > Today's Topics: > > 1. Re: Eudyptula challenge status (Drew Fustini) > 2. RE: lots of connections in SYN_RECV state (Puneet Agarwal) > 3. Re: lots of connections in SYN_RECV state > (Valdis.Kletnieks(a)vt.edu) > 4. Re: lots of connections in SYN_RECV state (Dave Tian) > 5. RE: lots of connections in SYN_RECV state (Puneet Agarwal) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Fri, 7 Nov 2014 11:04:28 -0600 > From: Drew Fustini <pdp7pdp7(a)gmail.com> > Subject: Re: Eudyptula challenge status > To: Dan <qsdconsulting(a)gmail.com> > Cc: kernelnewbies <kernelnewbies(a)kernelnewbies.org> > Message-ID: > < > CAEf4M_B6W3GJ8P_ShhpoL8QqtGEZ7h18p92BxuH0foQOQYj2Wg(a)mail.gmail.com> > Content-Type: text/plain; charset=UTF-8 > > Yup, I submitted my current task 4 weeks ago, sent a note last week, > and got reply the next day (Nov 1) with same sentiment: relax - > everything is ok, it's a slow process, and it's not a race. > > On Tue, Nov 4, 2014 at 8:59 PM, Dan <qsdconsulting(a)gmail.com> wrote: > > Ramon Fried <ramon.fried <at> tandemg.com> writes: > > > >> > >> > >> Hey all. > >> Tasks are pending for a long time. Anyone has Info regarding the queue? > >> Thanks. > >> Ramon > >> > >> > >> _______________________________________________ > >> Kernelnewbies mailing list > >> Kernelnewbies <at> kernelnewbies.org > >> http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies > >> > > > > > > I've been on challenge 5 for about a month and a half. I sent him a > follow > > up email a few weeks ago and he replied "Relax, things are slow, there's > no > > rush..." > > > > His response times before that were usually less than 2 days. > > > > > > > > _______________________________________________ > > Kernelnewbies mailing list > > Kernelnewbies(a)kernelnewbies.org > > http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies > > > > ------------------------------ > > Message: 2 > Date: Fri, 7 Nov 2014 23:11:26 +0530 > From: Puneet Agarwal <puneet.agr(a)outlook.com> > Subject: RE: lots of connections in SYN_RECV state > To: Dave Tian <dave.jing.tian(a)gmail.com>, Silvan Jegen > <me(a)sillymon.ch> > Cc: "kernelnewbies(a)kernelnewbies.org" > <kernelnewbies(a)kernelnewbies.org> > Message-ID: <SNT153-W7011B59E54B5FD99A3604699850(a)phx.gbl> > Content-Type: text/plain; charset="iso-8859-1" > > I use linux kernel 2.6. I have enabled SYN cookies already. But that does > not seem to solve the problem. Overall request latency is very high with > these many half open connections. > > Thanks and Regards > Puneet > ---------------------------------------- > > Subject: Re: lots of connections in SYN_RECV state > > From: dave.jing.tian(a)gmail.com > > Date: Fri, 7 Nov 2014 23:49:35 +0800 > > CC: puneet.agr(a)outlook.com; kernelnewbies(a)kernelnewbies.org > > To: me(a)sillymon.ch > > > > Latest kernel provides a TCP SYN Cookie feature to defense from SYN > flooding. > > > > -daveti > > > > > >> On Nov 6, 2014, at 11:58 PM, Silvan Jegen <me(a)sillymon.ch> wrote: > >> > >> 2014-11-06 16:15,Puneet Agarwal: > >>> Is there a way to check the reason, why they do not answer to the > >>> SYN-ACK's? > >> > >> I don't think so. After all, they just don't answer and they won't tell > >> you why (AFAIK there is no way to ask them why either)... > >> > >> You could try to check for patterns in the incoming IP addresses to see > >> from how many different places these connections are being made. I think > >> that way it should be possible to figure out from which geographic > >> location these problematic connections are coming from as well. What you > >> would do with these findings I am not sure though. > >> > >> If these connection negatively impact the performance of your servers > >> you should definitely look into to countermeasures mentioned in the RFC > >> here. > >> > >> http://tools.ietf.org/html/rfc4987 > >> > >> > >> _______________________________________________ > >> Kernelnewbies mailing list > >> Kernelnewbies(a)kernelnewbies.org > >> http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies > > > > > > ------------------------------ > > Message: 3 > Date: Fri, 07 Nov 2014 13:10:05 -0500 > From: Valdis.Kletnieks(a)vt.edu > Subject: Re: lots of connections in SYN_RECV state > To: Puneet Agarwal <puneet.agr(a)outlook.com> > Cc: Dave Tian <dave.jing.tian(a)gmail.com>, Silvan Jegen > <me(a)sillymon.ch>, "kernelnewbies(a)kernelnewbies.org" > <kernelnewbies(a)kernelnewbies.org> > Message-ID: <9736.1415383805(a)turing-police.cc.vt.edu> > Content-Type: text/plain; charset="us-ascii" > > On Fri, 07 Nov 2014 23:11:26 +0530, Puneet Agarwal said: > > > I use linux kernel 2.6. I have enabled SYN cookies already. But that > does not > > seem to solve the problem. Overall request latency is very high with > these many > > half open connections. > > So, out of curiosity, where are all these half open connections coming > from? Are they from addresses in your local network? Outside sites that > *should* be connecting? Places you've never heard and and probably > *shouldn't* > be connecting? > > (Also, if you have properly implemented syncookies, you shouldn't *have* > any > half-open connections. That's the whole point of syncookies....) > >

2 2