On Tue, Mar 22, 2016 at 8:06 PM, <kernelnewbies-request@xxxxxxxxxxxxxxxxx> wrote:
Send Kernelnewbies mailing list submissions to
kernelnewbies@xxxxxxxxxxxxxxxxx
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@xxxxxxxxxxxxxxxxx
You can reach the person managing the list at
kernelnewbies-owner@xxxxxxxxxxxxxxxxx
When replying, please edit your Subject line so it is more specific
than "Re: Contents of Kernelnewbies digest..."
Today's Topics:
1. Re: Is it possible to turn off the gcc optimization when
compiling kernel? (Hao Lee)
2. Re: kernel_thread() causes segfault (Shashank Khasare)
3. Re: Is it possible to turn off the gcc optimization when
compiling kernel? (Nicholas Mc Guire)
4. Re: How to get object virtual address from a kernel core dump
(Arun Sudhilal)
5. Re: Is it possible to turn off the gcc optimization when
compiling kernel? (Hao Lee)
6. Re: kernel_thread() causes segfault (Manoj Nayak)
7. Re :Querry Regarding Memory Alignment (Manoj Nayak)
----------------------------------------------------------------------
Message: 1
Date: Tue, 22 Mar 2016 19:24:38 +0800
From: Hao Lee <haolee.swjtu@xxxxxxxxx>
Subject: Re: Is it possible to turn off the gcc optimization when
compiling kernel?
To: kernelnewbies@xxxxxxxxxxxxxxxxx
Cc: Nicholas Mc Guire <der.herr@xxxxxxx>
Message-ID:
<CA+PpKPm=eHxK5oj39r8nrT6XGA4qo6XKMNRcgcNCNdGQW_SPvw@xxxxxxxxxxxxxx>
Content-Type: text/plain; charset=UTF-8
On Mon, Mar 21, 2016 at 5:51 PM, Nicholas Mc Guire <der.herr@xxxxxxx> wrote:
> You can not turn it off in all functions as some need particluar
> optimization flags to comile at all, but you can pass
> individual CFLAGS per file via the Makefile
>
> CFLAGS_target.o = -O0 or -flags-to-use
>
> aswell as remove specific CFLAGS with
>
> CFLAGS_REMOVE_target.o = -flags-to-remove
>
> but if you want to debug the kernel it is most likely not
> a good idea to try and disable optimization as the code you then
> are debugging might not have that much to do with the final code
> once optimization is on again. So simply generate the .lst file
> of the target you are trying to debug e.g. for kernel/sched/core.c:
>
> make kernel/sched/core.lst
>
> and then use that .lst file to understand the output of gdb you
> are inspecting.
Thanks for your reply!
Besides,I also find that use "gcc -c -Q -O1 --help=optimizers" can
print the exact set of optimizations.
regards,
Hao Lee
------------------------------
Message: 2
Date: Tue, 22 Mar 2016 16:51:44 +0530
From: Shashank Khasare <sskkernelnewbie@xxxxxxxxx>
Subject: Re: kernel_thread() causes segfault
To: kernelnewbies@xxxxxxxxxxxxxxxxx
Message-ID:
<CAGwMyOAxpQJjR9LRDLCp-ZLgHqKPcaLuJwDWyowBMsHHs5sjug@xxxxxxxxxxxxxx>
Content-Type: text/plain; charset="utf-8"
I am trying to implement ideas mentioned in the following OSDI paper:
L. Soares and M. Stumm. FlexSC: flexible system call scheduling with
exception-less system calls. In Proc. OSDI, 2010.
http://www.cs.cmu.edu/~chensm/Big_Data_reading_group/papers/flexsc-osdi10.pdf
The paper propose a new mechanism for applications to make syscall.
The brief idea is to have two types of threads 1) User thread 2) Kernel
Thread.
These two threads share same address space, file descriptor tables, parent
pid etc.
Whenever user thread wants to make syscall, it would post the information
about syscall number & arguments
to syscall in common shared page. User thread would then wait till the
results are posted on shared page.
The kernel thread reads the syscall arguments from shared page and writes
the results to shared page.
User thread consumes the results and continues execution.
Since the kernel thread and user thread can be scheduled on different cpu
cores, and user thread is ideally never executing
kernel code and vice versa, one can expect gain in instruction per cycle
for application, since the cache pollution is reduced to
some extent*.*
So to implement this mechanism, it is important for the user and kernel
thread to share address space, fd tables etc.
kernel_thread() works fine with older kernels to achieve this task, but is
no longer an option.
Is there of any mechanism for sharing fd tables as well?
Please let me know.
Thanks a lot,
Shashank
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.kernelnewbies.org/pipermail/kernelnewbies/attachments/20160322/1bdf0577/attachment-0001.html
------------------------------
Message: 3
Date: Tue, 22 Mar 2016 11:35:16 +0000
From: Nicholas Mc Guire <der.herr@xxxxxxx>
Subject: Re: Is it possible to turn off the gcc optimization when
compiling kernel?
To: Hao Lee <haolee.swjtu@xxxxxxxxx>
Cc: kernelnewbies@xxxxxxxxxxxxxxxxx
Message-ID: <20160322113516.GA11898@xxxxxxxx>
Content-Type: text/plain; charset=us-ascii
On Tue, Mar 22, 2016 at 07:24:38PM +0800, Hao Lee wrote:
> On Mon, Mar 21, 2016 at 5:51 PM, Nicholas Mc Guire <der.herr@xxxxxxx> wrote:
> > You can not turn it off in all functions as some need particluar
> > optimization flags to comile at all, but you can pass
> > individual CFLAGS per file via the Makefile
> >
> > CFLAGS_target.o = -O0 or -flags-to-use
> >
> > aswell as remove specific CFLAGS with
> >
> > CFLAGS_REMOVE_target.o = -flags-to-remove
> >
> > but if you want to debug the kernel it is most likely not
> > a good idea to try and disable optimization as the code you then
> > are debugging might not have that much to do with the final code
> > once optimization is on again. So simply generate the .lst file
> > of the target you are trying to debug e.g. for kernel/sched/core.c:
> >
> > make kernel/sched/core.lst
> >
> > and then use that .lst file to understand the output of gdb you
> > are inspecting.
>
> Thanks for your reply!
> Besides,I also find that use "gcc -c -Q -O1 --help=optimizers" can
> print the exact set of optimizations.
>
yes - but for any given code you will find that many of those
options actually have no effect for the particular code blob.
The set of flags effectively impacting the generated object
file is generally much smaller than the ones reported by
gcc -c -Q -O1 --help=optimizers.
thx!
hofrat
------------------------------
Message: 4
Date: Tue, 22 Mar 2016 17:09:08 +0530
From: Arun Sudhilal <getarunks@xxxxxxxxx>
Subject: Re: How to get object virtual address from a kernel core dump
To: "Mohammad Y. Zachariah" <eng.myz@xxxxxxxxx>
Cc: Kernel Newbies <kernelnewbies@xxxxxxxxxxxxxxxxx>
Message-ID:
<CABOM9ZrcecujZBfyLXSK8SXPFmkZGWRNH6Y-ZnuxQEZJVHJ15Q@xxxxxxxxxxxxxx>
Content-Type: text/plain; charset="utf-8"
Hello Zach,
On Fri, Mar 18, 2016 at 3:28 PM, Mohammad Y. Zachariah <eng.myz@xxxxxxxxx>
wrote:
> Hello everyone,
>
> I'm taking the way of analysing kernel core dumps as a learning approach
> using 'crash tool'. One of the interesting crash commands is 'struct' which
> can print kernel struct definition and/or the actual contents of the
> structure.
>
> According to struct help page, I need the virtual address of the struct in
> order to view/print its contents, for example:
>
> crash> mm_struct.pgd ffff810022e7d080 -px
> pgd_t *pgd = 0xffff81000e3ac000
> -> {
> pgd = 0x2c0a6067
> }
>
> My question is how to find the mm_struct address "ffff810022e7d080" in the
> above example in the first place??
>
crash tool has a 'ps' command, which outputs all the task and their task
struct address.
Thanks,
Arun
>
> Thank you for your help in advance.
> Zach
>
> _______________________________________________
> Kernelnewbies mailing list
> Kernelnewbies@xxxxxxxxxxxxxxxxx
> http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.kernelnewbies.org/pipermail/kernelnewbies/attachments/20160322/676f2c2d/attachment-0001.html
------------------------------
Message: 5
Date: Tue, 22 Mar 2016 19:46:23 +0800
From: Hao Lee <haolee.swjtu@xxxxxxxxx>
Subject: Re: Is it possible to turn off the gcc optimization when
compiling kernel?
To: Nicholas Mc Guire <der.herr@xxxxxxx>
Cc: kernelnewbies@xxxxxxxxxxxxxxxxx
Message-ID:
<CA+PpKPk4fEZqH7J9khjVoiduT3Sz+rBWpqpHBSk+BNVXLwF0eQ@xxxxxxxxxxxxxx>
Content-Type: text/plain; charset=UTF-8
On Tue, Mar 22, 2016 at 7:35 PM, Nicholas Mc Guire <der.herr@xxxxxxx> wrote:
> yes - but for any given code you will find that many of those
> options actually have no effect for the particular code blob.
> The set of flags effectively impacting the generated object
> file is generally much smaller than the ones reported by
> gcc -c -Q -O1 --help=optimizers.
>
> thx!
> hofrat
yeah,only several options take effect.
Because "-O0" will break the link progress,I just want use "-O1" and
some other options to approximate the effects of "-O0".
regards,
Hao Lee
------------------------------
Message: 6
Date: Tue, 22 Mar 2016 19:15:48 +0530
From: Manoj Nayak <manojnayak2005@xxxxxxxxx>
Subject: Re: kernel_thread() causes segfault
To: kernelnewbies@xxxxxxxxxxxxxxxxx
Message-ID:
<CAOsfsFLeO5v5X5obrUeokq=WLdQO3O=PjmbfYAU4WB334CBddw@xxxxxxxxxxxxxx>
Content-Type: text/plain; charset="utf-8"
Process has files_struct and fs_struct in task_struct.
Two thread's task_struct can point to same files_struct and fs_struct if
we do the changes through a new system call.
Please check the following URL.
http://lxr.free-electrons.com/source/kernel/fork.c#L993
Regards
Manoj Nayak
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.kernelnewbies.org/pipermail/kernelnewbies/attachments/20160322/687ef0ce/attachment-0001.html
------------------------------
Message: 7
Date: Tue, 22 Mar 2016 20:00:05 +0530
From: Manoj Nayak <manojnayak2005@xxxxxxxxx>
Subject: Re :Querry Regarding Memory Alignment
To: kernelnewbies@xxxxxxxxxxxxxxxxx
Message-ID:
<CAOsfsFKdyWR6t6k7WKPwHEe_zRz9KwiUZvpJQCOCqRgOmvx39w@xxxxxxxxxxxxxx>
Content-Type: text/plain; charset="utf-8"
The CPU always reads at its word size (4 bytes on a 32-bit processor), so
when you do an unaligned address access ? on a processor that supports it ?
the processor is going to read multiple words. The CPU will read each word
of memory that your requested address straddles. This causes multiple
processor read to access the requested data.
Hardware deals with word size alignment. Other alignment like page size
alignment is a software feature.
On X86: Any virtual address provided to cpu is split into page aligned
virtual address and an offset.CPU uses page aligned virtual address to
check TLB to get page aligned physical address.
Now physical address = Page aligned physical address + offset
If page size is bigger then less no of TLB entries are required for the
processor. Each page needs one TLB entry.
Regards
Manoj Nayak
Thanks Manoj for the reply
So that means aligned / unaligned virtual address -------- > aligned/unaligned phy address
and there is one to one analogy in this regard.
SW can actually generate addresses which could be unaligned and then CPU could fault with
memory alignment fault.
For example, if SW cast a "short" pointer to a "int" pointer and attempts to access the data
it leads to unaligned access.
Perhaps the Instruction decode unit of the CPU will see that it is word read at address which is not divisible by 4
and would instruct the control unit to trigger the alignment fault.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.kernelnewbies.org/pipermail/kernelnewbies/attachments/20160322/56d79c72/attachment.html
------------------------------
_______________________________________________
Kernelnewbies mailing list
Kernelnewbies@xxxxxxxxxxxxxxxxx
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
End of Kernelnewbies Digest, Vol 64, Issue 37
*********************************************
_______________________________________________ Kernelnewbies mailing list Kernelnewbies@xxxxxxxxxxxxxxxxx http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
