[PATCH v5 00/15] target/ppc: DFP instructions using decodetree

Luis Pires posted 15 patches 2 years, 6 months ago
Test checkpatch failed
Patches applied successfully (tree, apply log)
git fetch https://github.com/patchew-project/qemu tags/patchew/20211029192417.400707-1-luis.pires@eldorado.org.br
Maintainers: Greg Kurz <groug@kaod.org>, David Gibson <david@gibson.dropbear.id.au>
include/libdecnumber/decNumber.h       |   4 +
include/libdecnumber/decNumberLocal.h  |   2 +-
include/qemu/host-utils.h              |  36 +++
libdecnumber/decContext.c              |   7 +-
libdecnumber/decNumber.c               | 131 ++++++++
target/ppc/dfp_helper.c                | 168 +++++++---
target/ppc/helper.h                    | 106 ++++---
target/ppc/insn32.decode               | 171 ++++++++++
target/ppc/translate.c                 |  23 +-
target/ppc/translate/dfp-impl.c.inc    | 419 ++++++++++++-------------
target/ppc/translate/dfp-ops.c.inc     | 165 ----------
target/ppc/translate/vector-impl.c.inc |  10 +-
12 files changed, 747 insertions(+), 495 deletions(-)
delete mode 100644 target/ppc/translate/dfp-ops.c.inc
[PATCH v5 00/15] target/ppc: DFP instructions using decodetree
Posted by Luis Pires 2 years, 6 months ago
This series moves all existing DFP instructions to decodetree and
implements the 2 new instructions (dcffixqq and dctfixqq) from
Power ISA 3.1.

In order to implement dcffixqq, divu128/divs128 were modified to
support 128-bit quotients (previously, they were limited to 64-bit
quotients), along with adjustments being made to their existing callers.
libdecnumber was also expanded to allow creating decimal numbers from
128-bit integers.

Similarly, for dctfixqq, mulu128 (host-utils) and decNumberIntegralToInt128
(libdecnumber) were introduced to support 128-bit integers.

The remaining patches of this series move all of the already existing
DFP instructions to decodetree, and end up removing dfp-ops.c.inc, which
is no longer needed.

---

I'm resending this series because I missed adding a R-b tag on one
of the commits in v4. I also dropped patches that were already applied and
rebased against master.

All of these patches are reviewed and ready to be applied.

Changes in v5:
- Rebased against master
- Skipped patches that were already applied (v4 1-4)

Changes in v4:
- Rebased against dgibson/ppc-for-6.2
- Skipped patches that were already applied (v3 1-4)
- Changed divu128/divs128 to return the remainder (rth)
- Moved changes that drop nip updates to a separate patch (rth)

Changes in v3:
- Split the uabs64 patch in 2
- Included patch to fix missing zero-extension in divs128
- Folded divisor == 0 into the dhi == 0 case in divu128
- Moved udiv_qrnnd from softfloat-macros.h to host-utils.h
- Used udiv_qrnnd in divu128
- Replaced int with bool in divs128
- Added unit test to check the divisor normalization in divu128
- Removed 'inline' from times_* functions in ppc/translate.c
- Used uadd64_overflow in mulu128
- Removed unnecessary 'else' from decNumberIntegralToInt128

Changes in v2:
- Renamed abs64() to uabs64()

Bruno Larsen (1):
  target/ppc: Move REQUIRE_ALTIVEC/VECTOR to translate.c

Fernando Valle (1):
  target/ppc: Introduce REQUIRE_FPU

Luis Pires (13):
  libdecnumber: introduce decNumberFrom[U]Int128
  target/ppc: Implement DCFFIXQQ
  host-utils: Introduce mulu128
  libdecnumber: Introduce decNumberIntegralToInt128
  target/ppc: Implement DCTFIXQQ
  target/ppc: Do not update nip on DFP instructions
  target/ppc: Move dtstdc[q]/dtstdg[q] to decodetree
  target/ppc: Move d{add,sub,mul,div,iex}[q] to decodetree
  target/ppc: Move dcmp{u,o}[q],dts{tex,tsf,tsfi}[q] to decodetree
  target/ppc: Move dquai[q], drint{x,n}[q] to decodetree
  target/ppc: Move dqua[q], drrnd[q] to decodetree
  target/ppc: Move dct{dp,qpq},dr{sp,dpq},dc{f,t}fix[q],dxex[q] to
    decodetree
  target/ppc: Move ddedpd[q],denbcd[q],dscli[q],dscri[q] to decodetree

 include/libdecnumber/decNumber.h       |   4 +
 include/libdecnumber/decNumberLocal.h  |   2 +-
 include/qemu/host-utils.h              |  36 +++
 libdecnumber/decContext.c              |   7 +-
 libdecnumber/decNumber.c               | 131 ++++++++
 target/ppc/dfp_helper.c                | 168 +++++++---
 target/ppc/helper.h                    | 106 ++++---
 target/ppc/insn32.decode               | 171 ++++++++++
 target/ppc/translate.c                 |  23 +-
 target/ppc/translate/dfp-impl.c.inc    | 419 ++++++++++++-------------
 target/ppc/translate/dfp-ops.c.inc     | 165 ----------
 target/ppc/translate/vector-impl.c.inc |  10 +-
 12 files changed, 747 insertions(+), 495 deletions(-)
 delete mode 100644 target/ppc/translate/dfp-ops.c.inc

-- 
2.25.1


Re: [PATCH v5 00/15] target/ppc: DFP instructions using decodetree
Posted by David Gibson 2 years, 5 months ago
On Fri, Oct 29, 2021 at 04:24:02PM -0300, Luis Pires wrote:
> This series moves all existing DFP instructions to decodetree and
> implements the 2 new instructions (dcffixqq and dctfixqq) from
> Power ISA 3.1.
> 
> In order to implement dcffixqq, divu128/divs128 were modified to
> support 128-bit quotients (previously, they were limited to 64-bit
> quotients), along with adjustments being made to their existing callers.
> libdecnumber was also expanded to allow creating decimal numbers from
> 128-bit integers.
> 
> Similarly, for dctfixqq, mulu128 (host-utils) and decNumberIntegralToInt128
> (libdecnumber) were introduced to support 128-bit integers.
> 
> The remaining patches of this series move all of the already existing
> DFP instructions to decodetree, and end up removing dfp-ops.c.inc, which
> is no longer needed.

Applied to ppc-for-6.2, thanks.

> 
> ---
> 
> I'm resending this series because I missed adding a R-b tag on one
> of the commits in v4. I also dropped patches that were already applied and
> rebased against master.
> 
> All of these patches are reviewed and ready to be applied.
> 
> Changes in v5:
> - Rebased against master
> - Skipped patches that were already applied (v4 1-4)
> 
> Changes in v4:
> - Rebased against dgibson/ppc-for-6.2
> - Skipped patches that were already applied (v3 1-4)
> - Changed divu128/divs128 to return the remainder (rth)
> - Moved changes that drop nip updates to a separate patch (rth)
> 
> Changes in v3:
> - Split the uabs64 patch in 2
> - Included patch to fix missing zero-extension in divs128
> - Folded divisor == 0 into the dhi == 0 case in divu128
> - Moved udiv_qrnnd from softfloat-macros.h to host-utils.h
> - Used udiv_qrnnd in divu128
> - Replaced int with bool in divs128
> - Added unit test to check the divisor normalization in divu128
> - Removed 'inline' from times_* functions in ppc/translate.c
> - Used uadd64_overflow in mulu128
> - Removed unnecessary 'else' from decNumberIntegralToInt128
> 
> Changes in v2:
> - Renamed abs64() to uabs64()
> 
> Bruno Larsen (1):
>   target/ppc: Move REQUIRE_ALTIVEC/VECTOR to translate.c
> 
> Fernando Valle (1):
>   target/ppc: Introduce REQUIRE_FPU
> 
> Luis Pires (13):
>   libdecnumber: introduce decNumberFrom[U]Int128
>   target/ppc: Implement DCFFIXQQ
>   host-utils: Introduce mulu128
>   libdecnumber: Introduce decNumberIntegralToInt128
>   target/ppc: Implement DCTFIXQQ
>   target/ppc: Do not update nip on DFP instructions
>   target/ppc: Move dtstdc[q]/dtstdg[q] to decodetree
>   target/ppc: Move d{add,sub,mul,div,iex}[q] to decodetree
>   target/ppc: Move dcmp{u,o}[q],dts{tex,tsf,tsfi}[q] to decodetree
>   target/ppc: Move dquai[q], drint{x,n}[q] to decodetree
>   target/ppc: Move dqua[q], drrnd[q] to decodetree
>   target/ppc: Move dct{dp,qpq},dr{sp,dpq},dc{f,t}fix[q],dxex[q] to
>     decodetree
>   target/ppc: Move ddedpd[q],denbcd[q],dscli[q],dscri[q] to decodetree
> 
>  include/libdecnumber/decNumber.h       |   4 +
>  include/libdecnumber/decNumberLocal.h  |   2 +-
>  include/qemu/host-utils.h              |  36 +++
>  libdecnumber/decContext.c              |   7 +-
>  libdecnumber/decNumber.c               | 131 ++++++++
>  target/ppc/dfp_helper.c                | 168 +++++++---
>  target/ppc/helper.h                    | 106 ++++---
>  target/ppc/insn32.decode               | 171 ++++++++++
>  target/ppc/translate.c                 |  23 +-
>  target/ppc/translate/dfp-impl.c.inc    | 419 ++++++++++++-------------
>  target/ppc/translate/dfp-ops.c.inc     | 165 ----------
>  target/ppc/translate/vector-impl.c.inc |  10 +-
>  12 files changed, 747 insertions(+), 495 deletions(-)
>  delete mode 100644 target/ppc/translate/dfp-ops.c.inc
> 

-- 
David Gibson			| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you.  NOT _the_ _other_
				| _way_ _around_!
http://www.ozlabs.org/~dgibson