floating points in kernel space
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*
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@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
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
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
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
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
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
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@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
participants (7)
-
Dave Hylands -
Greg Freemyer -
julie Sullivan -
lei yang -
mohit verma -
Mulyadi Santosa -
Victor Rodriguez