Re: [Qemu-devel] [PATCH 1/2] softfloat: Enable run-time-configurable meaning of signaling NaN bit

[Bastian Koppelmann <kbastian@mail.uni-paderborn.de>]

On 03/25/2016 01:50 PM, Aleksandar Markovic wrote:
> From: Aleksandar Markovic <aleksandar.markovic@imgtec.com>
> 
> This patch enables SoftFloat library to be configured at run-time in
> relation to the meaning of signaling NaN bit.
> 
> Background:
> 
> In floating-point calculations, there is a need for denoting undefined or
> unrepresentable values. This is achieved by defining certain floating-point
> numerical values to be NaNs (which stands for "not a number"). For additional
> reasons, virtually all modern floating-point unit implementations use two
> kinds of NaNs: quiet and signaling. The binary representations of these two
> kinds of NaNs, as a rule, differ only in one bit (it is traditionally,
> the first bit of mantissa).
> 
> Up to 2008, standards for floating-point did not specify all details about
> binary representation of NaNs. More specifically, the meaning of the bit
> that is used for distinguishing between signaling and quiet NaNs was not
> strictly prescribed. (IEEE 754-2008 was the first floating-point standard
> that defined that meaning clearly, see [1], p. 35) As a result, different
> platforms took different approaches, and this presented certain challenge
> in emulators like QEMU.
> 
> Mips architecture represent the most complex case among QEMU-supported
> architectures regarding signaling NaN bit. Up to Release 6 of Mips
> architecture, "1" in signaling NaN bit denoted signaling NaN, which is
> opposite to IEEE 754-2008 standard. From Release 6 on, Mips architecture
> adopted IEEE standard prescription, and "0" denotes signaling NaN. On top of
> that, Mips architecture for SIMD (also known as MSA, or vector instructions)
> also specifies signaling bit in accordance to IEEE standard. MSA unit can be
> implemented with both pre-Release 6 and Release 6 main processor units.
> 
> QEMU uses SoftFloat library to implement various floating-point-related
> instructions on all platforms. The current implementation allows for defining
> meaning of signaling NaN bit during build time, and is implemented via
> preprocessor macro called SNAN_BIT_IS_ONE.
> 
> The change in this patch enables SoftFloat library to be configured in
> run-time. This configuration is meant to occur during CPU initialization,
> when it is definitely known what desired behavior for particular CPU
> (or any additional FPUs) is.
> 
> The change is implemented so that it is consistent with existing
> implementation of similar cases. This means that structure float_status is
> used for passing the information about desired signaling NaN bit during each
> invocation of SoftFloat functions. The additional field in float_status is
> called snan_bit_is_one, which supersedes macro SNAN_BIT_IS_ONE.
> 
> Further break down of changes:
> 
>   (for the sake of brevity, a placeholder XXX is used below and it might
>   mean float16, float32, float64, floatx80, or float128)
> 
>   1) Added field snan_bit_is_one to the structure float_status,
>      and the correspondent setter function set_snan_bit_is_one().
> 
>   2) SoftFloat library constants XXX_default_nan converted to functions
>      XXX_default_nan(float_status*). This is necessary since they are
>      dependant on signaling bit meaning.
> 
>   3) Added a float_status* argument to SoftFloat library functions
>      XXX_is_quiet_nan(XXX a_), XXX_is_signaling_nan(XXX a_),
>      XXX_maybe_silence_nan(XXX a_).
> 
>   4) Updated code in all architectures to reflect changes in SoftFloat
>      library. This change is twofolds: it includes modification of SoftFloat
>      library functions invocations, and addition of invocations of function
>      set_snan_bit_is_one() during CPU initialization, with arguments that
>      are appropriate for each architecture.
> 
> IMPORTANT:
> 
> This change is not meant to create any change in emulator behavior or
> functionality on any platform. It just provides the means for SoftFloat
> library to be used in a more flexible way - in other words, it will just
> prepare SoftFloat library for usage related to Mips platform and its
> specifics regarding signaling bit meaning, which is done in the next patch.
> 
> [1] "IEEE Standard for Floating-Point Arithmetic",
>     IEEE Computer Society, August 29, 2008.
>     (http://www.csee.umbc.edu/~tsimo1/CMSC455/IEEE-754-2008.pdf)
> 
> Signed-off-by: Aleksandar Markovic <aleksandar.markovic@imgtec.com>
> ---
>  fpu/softfloat-specialize.h    | 546 +++++++++++++++++++++---------------------
>  fpu/softfloat.c               | 170 +++++++------
>  include/fpu/softfloat.h       |  49 ++--
>  target-alpha/cpu.c            |   2 +
>  target-arm/cpu.c              |   2 +
>  target-arm/helper-a64.c       |  14 +-
>  target-arm/helper.c           |  40 ++--
>  target-i386/cpu.c             |   4 +
>  target-m68k/cpu.c             |   2 +
>  target-m68k/helper.c          |   6 +-
>  target-microblaze/cpu.c       |   2 +
>  target-microblaze/op_helper.c |   6 +-
>  target-mips/helper.h          |   4 +-
>  target-mips/msa_helper.c      |  96 ++++----
>  target-mips/op_helper.c       |  13 +-
>  target-mips/translate.c       |   4 +-
>  target-mips/translate_init.c  |   6 +-
>  target-openrisc/cpu.c         |   2 +
>  target-ppc/fpu_helper.c       | 120 +++++-----
>  target-ppc/translate_init.c   |   2 +
>  target-s390x/cpu.c            |   1 +
>  target-s390x/fpu_helper.c     |  28 ++-
>  target-s390x/helper.h         |   6 +-
>  target-s390x/translate.c      |   6 +-
>  target-sh4/cpu.c              |   1 +
>  target-sparc/cpu.c            |   1 +
>  target-tricore/helper.c       |   1 +
>  target-unicore32/cpu.c        |   1 +
>  target-xtensa/cpu.c           |   3 +
>  29 files changed, 608 insertions(+), 530 deletions(-)
> 

For the TriCore part you can have a

Tested-by: Bastian Koppelmann<kbastian@mail.uni-paderborn.de>

Cheers,
Bastian