Commits

Breadcrumbs

History for

one99c678

User selector

All users

Datepicker

All time

Commit History

Commits on Dec 18, 2005

• Major 64-bit cleanup.

  matteo-frigo

  committed
  e99c678

  Copy full SHA for e99c678

Commits on Sep 27, 2005

• Fixed wrong opcount for simd codelets.

  matteo-frigo

  committed
  2c35b6d

  Copy full SHA for 2c35b6d

• Implemented split-complex SIMD codelets

  matteo-frigo

  committed
  804b1a4

  Copy full SHA for 804b1a4

Commits on Sep 26, 2005

• Generalized the store pairs'' trick (now called store multiple'').

  matteo-frigo

  committed
  4c34b95

  Copy full SHA for 4c34b95

Commits on Sep 11, 2005

• Higher size limit for t2 codelets.

  matteo-frigo

  committed
  5750c65

  Copy full SHA for 5750c65

• Heuristic: do not use t2 simd codelets for N>1024.

  matteo-frigo

  committed
  c5134ff

  Copy full SHA for c5134ff

Commits on Apr 10, 2005

• joned L-U-planner branch

  matteo-frigo

  committed
  cc67338

  Copy full SHA for cc67338

Commits on Mar 20, 2005

• Enable vector recursion for in-place problems, otherwise

  Show description for 094cbe9
  matteo-frigo

  committed
  094cbe9

  Copy full SHA for 094cbe9

• oops

  matteo-frigo

  committed
  14a9b59

  Copy full SHA for 14a9b59

• make solver UGLY for small N

  matteo-frigo

  committed
  7ea889c

  Copy full SHA for 7ea889c

• new dftw-genericbuf solver

  matteo-frigo

  committed
  a4abcfa

  Copy full SHA for a4abcfa

• new dftw-genericbuf solver

  matteo-frigo

  committed
  70997fb

  Copy full SHA for 70997fb

Commits on Mar 15, 2005

• Added genfft flag -precompute-twiddles which moves the computation of

  Show description for 155f07c
  matteo-frigo

  committed
  155f07c

  Copy full SHA for 155f07c

Commits on Mar 12, 2005

• Do not approximate pcost = vl * child->pcost unless child is guaranteed

  Show description for 5569655
  matteo-frigo

  committed
  5569655

  Copy full SHA for 5569655

Commits on Mar 10, 2005

• Relaxed applicability conditions.

  matteo-frigo

  committed
  a5282a5

  Copy full SHA for a5282a5

Commits on Mar 9, 2005

• Minor optimization

  matteo-frigo

  committed
  2496640

  Copy full SHA for 2496640

• Paranoia

  matteo-frigo

  committed
  55b8abd

  Copy full SHA for 55b8abd

Commits on Mar 7, 2005

• Do not try radix-2 generic.

  matteo-frigo

  committed
  10a57b3

  Copy full SHA for 10a57b3

Commits on Mar 6, 2005

• Runtime checks to guarantee small strides.

  matteo-frigo

  committed
  562882d

  Copy full SHA for 562882d

• Reduced the search space for rank-0 transforms

  matteo-frigo

  committed
  1fcf241

  Copy full SHA for 1fcf241

Commits on Mar 1, 2005

• Implemented directbuf, enabled for now.

  matteo-frigo

  committed
  495b9d7

  Copy full SHA for 495b9d7

• Unified dftw-direct, dftw-directbuf in an attempt to tame code

  Show description for 1869b02
  matteo-frigo

  committed
  1869b02

  Copy full SHA for 1869b02

Commits on Feb 27, 2005

• Transposed the buffer, and skewed it. This allows for contiguous

  Show description for 9c7a7d3
  matteo-frigo

  committed
  9c7a7d3

  Copy full SHA for 9c7a7d3

Commits on Feb 25, 2005

• Because of the recent changes to kernel/pickdim.c, splitrnk=0 is no

  Show description for 4350026
  stevengj

  committed
  4350026

  Copy full SHA for 4350026

• tweak

  stevengj

  committed
  3bfeb64

  Copy full SHA for 3bfeb64

• slight relaxation

  stevengj

  committed
  24560b2

  Copy full SHA for 24560b2

• cruft

  stevengj

  committed
  cadf7b9

  Copy full SHA for cadf7b9

• added experimental indirect-transpose solver: when transforming the columns of the matrix, allow us to do a transpose to make the DFTs contiguous

  stevengj

  committed
  42d46a1

  Copy full SHA for 42d46a1

• require finite_rnk

  stevengj

  committed
  47e79fc

  Copy full SHA for 47e79fc

Commits on Feb 23, 2005

• Put cpy2d_pair into its own file, so that I can experiment with

  Show description for 5844ac6
  matteo-frigo

  committed
  5844ac6

  Copy full SHA for 5844ac6

Commits on Feb 19, 2005

• Make the batch size B=Theta(r) instead of B=Theta(1) in buffered

  Show description for 7709525
  matteo-frigo

  committed
  7709525

  Copy full SHA for 7709525

Commits on Feb 16, 2005

• Do not use SIMD_CFLAGS. The theory is that if taint.c is unsafe

  Show description for 79f7093
  matteo-frigo

  committed
  79f7093

  Copy full SHA for 79f7093

• Do not override CFLAGS in Makefile.am.

  matteo-frigo

  committed
  f9e6da5

  Copy full SHA for f9e6da5

Commits on Feb 14, 2005

• unified fma and non-fma versions

  matteo-frigo

  committed
  485e6db

  Copy full SHA for 485e6db

• Merged dft/codelets/inplace with the main dft/codelets/standard

  Show description for 48bfe71
  matteo-frigo

  committed
  48bfe71

  Copy full SHA for 48bfe71

Pagination

Previous

Next