floating points in kernel space

floating points in kernel space

[mohit verma]

hi all,
i read somewhere that there is no floating point arithmatic supported in the
kernel space . is it true??
but the floating point registers and even floating point arithmatic units
are all handled by the.........( kernel i think).


thanks in advance for help...........


-- 
........................
*MOHIT VERMA*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.kernelnewbies.org/pipermail/kernelnewbies/attachments/20110104/069a6cc2/attachment.html 

floating points in kernel space

[Victor Rodriguez]

On Tue, Jan 4, 2011 at 6:43 AM, mohit verma <mohit89mlnc@gmail.com> wrote:
> hi all,
> i read somewhere that there is no floating point arithmatic supported in the
> kernel space . is it true??
> but the floating point registers and even floating point arithmatic units
> are all handled by the.........( kernel i think).
>
>
> thanks in advance for help...........
>


HI it is true , the floating point is not supported on printk , AFAIK
the only way to handle is to send it like HEX values and then some
other application like perl or python transform it to floating point

Hope it helps

Regards

Victor Rodriguez
> --
> ........................
> MOHIT VERMA
>
> _______________________________________________
> Kernelnewbies mailing list
> Kernelnewbies at kernelnewbies.org
> http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
>
>

floating points in kernel space

[Dave Hylands]

Hi Mohit,

On Tue, Jan 4, 2011 at 4:43 AM, mohit verma <mohit89mlnc@gmail.com> wrote:
> hi all,
> i read somewhere that there is no floating point arithmatic supported in the
> kernel space . is it true??
> but the floating point registers and even floating point arithmatic units
> are all handled by the.........( kernel i think).

That is correct. In some architectures, attempts to use floating point
from the kernel will work. I've seen some x86 code that uses it.
However, with ARM for example, there is no float support in the
kernel, and some ARM architectures have no floating point support in
the hardware either. For ARM, there is a kernel implemented emulation
of the floating point instructions, but these can only be called from
user space. There are also some ARM software floating point libraries
(aka soft-fp) which can be used from user space (which are much more
efficient than using the kernel emulation library)

The kernel also doesn't support 64-bit division using the C divide
statement, unless you happen to be running on a 64-bit architecture.
There are some helper functions to achieve 64-bit division (see
div64.h).

Dave Hylands

floating points in kernel space

[Mulyadi Santosa]

Hi..

On Tue, Jan 4, 2011 at 22:44, Dave Hylands <dhylands@gmail.com> wrote:
> That is correct. In some architectures, attempts to use floating point
> from the kernel will work. I've seen some x86 code that uses it.


AFAIK, once x86 didn't supported due to floating point related
registers are not correctly (or even doing?) saved and restored during
context switching. So maybe it is fixed now...

-- 
regards,

Mulyadi Santosa
Freelance Linux trainer and consultant

blog: the-hydra.blogspot.com
training: mulyaditraining.blogspot.com

floating points in kernel space

[julie Sullivan]

>
> On Tue, Jan 4, 2011 at 22:44, Dave Hylands <dhylands@gmail.com> wrote:
> > That is correct. In some architectures, attempts to use floating point
> > from the kernel will work. I've seen some x86 code that uses it.
>
>
> AFAIK, once x86 didn't supported due to floating point related
> registers are not correctly (or even doing?) saved and restored during
> context switching. So maybe it is fixed now...
>
> --
>

I've often wondered about this oft-cited kernel behaviour too, in my
naivety.  I understand
that this must be on a per-arch basis, but does this mean that the kernel
doesn't police FP
access at _all_ (perhaps this is what Mohit means too)? Does code like X for
example have to access it directly, or does it just use the GPU? What about
other user-space code - does it
have a separate library and do its own security? Video drivers?
Sorry if these are basic questions, I grepped for float in the kernel but
as-yet the associated
code looks really arcane to me - if anyone could answer any of these
questions generally
(if that's possible) that would be very helpful with visualizing the
mechanism.
Maybe I'm looking in the wrong place.

When I started looking at the kernel I imagined this small, neat, concise
piece of highly
efficient code so I wasn't surprised there was no float (don't laugh - how
one learns :-/ ) ...
I suppose any float per-arch 'hacks' (to get a larger word size) would not
be worth
the overhead of the mode switch and extra code?

Thanks
Julie
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.kernelnewbies.org/pipermail/kernelnewbies/attachments/20110104/bb73a59d/attachment-0001.html 

floating points in kernel space

[Greg Freemyer]

On Tue, Jan 4, 2011 at 5:19 PM, julie Sullivan <kernelmail.jms@gmail.com> wrote:
>
>>
>> On Tue, Jan 4, 2011 at 22:44, Dave Hylands <dhylands@gmail.com> wrote:
>> > That is correct. In some architectures, attempts to use floating point
>> > from the kernel will work. I've seen some x86 code that uses it.
>>
>>
>> AFAIK, once x86 didn't supported due to floating point related
>> registers are not correctly (or even doing?) saved and restored during
>> context switching. So maybe it is fixed now...
>>
>> --
>
> I've often wondered about this oft-cited kernel behaviour too, in my
> naivety.? I understand
> that this must be on a per-arch basis, but does this mean that the kernel
> doesn't police FP
> access at _all_ (perhaps this is what Mohit means too)? Does code like X for
> example have to access it directly, or does it just use the GPU? What about
> other user-space code - does it
> have a separate library and do its own security? Video drivers?
> Sorry if these are basic questions, I grepped for float in the kernel but
> as-yet the associated
> code looks really arcane to me - if anyone could answer any of these
> questions generally
> (if that's possible) that would be very helpful with visualizing the
> mechanism.
> Maybe I'm looking in the wrong place.
>
> When I started looking at the kernel I imagined this small, neat, concise
> piece of highly
> efficient code so I wasn't surprised there was no float (don't laugh - how
> one learns :-/ ) ...
> I suppose any float per-arch 'hacks' (to get a larger word size) would not
> be worth
> the overhead of the mode switch and extra code?
>
> Thanks
> Julie

Julie,

I think the issue is the kernel is extremely concerned with the
efficiency of the syscall path.

Very legitimately some benchmarks just measure that one path to see
how many thousands of syscalls per second can be made.

To accelerate that path as much as possible, the linux kenel chooses
not to incur the overhead of preserving the FP registers on every
syscall.

So kernel code that uses FP must first ensure any registers it uses
are preserved.  I don't recall ever writing any FP kernel code, so I
don't know what facilities are available to do that.

Greg

floating points in kernel space

[julie Sullivan]

> I think the issue is the kernel is extremely concerned with the
> efficiency of the syscall path.
>
> Very legitimately some benchmarks just measure that one path to see
> how many thousands of syscalls per second can be made.
>
> To accelerate that path as much as possible, the linux kenel chooses
> not to incur the overhead of preserving the FP registers on every
> syscall.
>
> So kernel code that uses FP must first ensure any registers it uses
> are preserved.  I don't recall ever writing any FP kernel code, so I
> don't know what facilities are available to do that.
>
> Greg
>

Greg

I had a vague idea about overhead being incurred due to the mode switch, but

your clear explanation makes it a lot easier to understand _where_ this
happens.
It's very helpful, thank you for your response.

Julie
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.kernelnewbies.org/pipermail/kernelnewbies/attachments/20110104/ff816de8/attachment.html 

floating points in kernel space

[lei yang]

How can I know arm kernel has suppored hardware float point?
is there a kernel config to be enabled for this supporting?

Lei

On Wed, Jan 5, 2011 at 7:26 AM, julie Sullivan <kernelmail.jms@gmail.com> wrote:
>
>> I think the issue is the kernel is extremely concerned with the
>> efficiency of the syscall path.
>>
>> Very legitimately some benchmarks just measure that one path to see
>> how many thousands of syscalls per second can be made.
>>
>> To accelerate that path as much as possible, the linux kenel chooses
>> not to incur the overhead of preserving the FP registers on every
>> syscall.
>>
>> So kernel code that uses FP must first ensure any registers it uses
>> are preserved. ?I don't recall ever writing any FP kernel code, so I
>> don't know what facilities are available to do that.
>>
>> Greg
>
> Greg
>
> I had a vague idea about overhead being incurred due to the mode switch, but
> your clear explanation makes it a lot easier to understand _where_ this
> happens.
> It's very helpful, thank you for your response.
>
> Julie
>
>
> _______________________________________________
> Kernelnewbies mailing list
> Kernelnewbies at kernelnewbies.org
> http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
>
>