Commits

Branch selector

 56c7d29

User selector

All users

Datepicker

All time

Commit History

Commits on Mar 10, 2007

• Adopted MIPS_PS patches from Codesourcery.

  Show description for 56c7d29
  matteo-frigo

  committed
  56c7d29

  Copy full SHA for 56c7d29

Commits on Feb 25, 2007

• Incorrect initialization of win32 semaphores

  Show description for 117c18b
  matteo-frigo

  committed
  117c18b

  Copy full SHA for 117c18b

Commits on Jan 30, 2007

• win32 fixes (I think, still untested)

  Show description for 835fb99
  stevengj

  committed
  835fb99

  Copy full SHA for 835fb99

• Threading layer for Win32, completely untested.

  Show description for c3f9a60
  matteo-frigo

  committed
  c3f9a60

  Copy full SHA for c3f9a60

Commits on Jan 29, 2007

• Check for EINTR after sem_wait(), as suggested by Chip Salzenberg.

  Show description for e9103c2
  matteo-frigo

  committed
  e9103c2

  Copy full SHA for e9103c2

Commits on Jan 22, 2007

• Force vector recursion by means of a separate function pointer. I need this for Cell.

  Show description for 719f223
  matteo-frigo

  committed
  719f223

  Copy full SHA for 719f223

• Merge multiplications by twiddle with multiplications by i for faster r2c transforms.

  Show description for 44d6228
  matteo-frigo

  committed
  44d6228

  Copy full SHA for 44d6228

• Disabled vector recursion, too messy.

  Show description for 37defea
  matteo-frigo

  committed
  37defea

  Copy full SHA for 37defea

Commits on Jan 21, 2007

• Changed heuristics for vector recursion.

  Show description for f6f7ab5
  matteo-frigo

  committed
  f6f7ab5

  Copy full SHA for f6f7ab5

• Smarter algorithm for selection of nbuf.

  Show description for d49ea8d
  matteo-frigo

  committed
  d49ea8d

  Copy full SHA for d49ea8d

• Increased buffer sizes according to Moore's law.

  Show description for 983a3c8
  matteo-frigo

  committed
  983a3c8

  Copy full SHA for 983a3c8

Commits on Jan 19, 2007

• message-size heuristic in tranpose-recurse

  Show description for 22544bf
  stevengj

  committed
  22544bf

  Copy full SHA for 22544bf

• fix another MPI synchronization bug -- several more places where cost_hook must be called to synchronized process timings (sigh)

  Show description for 3063d37
  stevengj

  committed
  3063d37

  Copy full SHA for 3063d37

• Set havewisdom=0 when calling forget_wisdom() in the test program.

  Show description for ed26acb
  matteo-frigo

  committed
  ed26acb

  Copy full SHA for ed26acb

• remove redundant check

  Show description for 9c8fc20
  stevengj

  committed
  9c8fc20

  Copy full SHA for 9c8fc20

• fixed potential (unlikely) bug in wisdom import (triggered when importing impatient wisdom after creating more patient plans, but apparently only for nonstandard configure.c configurations)

  Show description for ea70911
  stevengj

  committed
  ea70911

  Copy full SHA for ea70911

• added functions to gather/broadcast wisdom for MPI

  Show description for c30ae9a
  stevengj

  committed
  c30ae9a

  Copy full SHA for c30ae9a

Commits on Jan 11, 2007

• whoops, another int/INT bug

  Show description for a87ad41
  stevengj

  committed
  a87ad41

  Copy full SHA for a87ad41

• whoops, fixed bug in transpose-recurse for r != m

  Show description for 51ddf45
  stevengj

  committed
  51ddf45

  Copy full SHA for 51ddf45

• canonicalize mpi-transposed flags by setting TRANSPOSED_IN/OUT where possible

  Show description for d255092
  stevengj

  committed
  d255092

  Copy full SHA for d255092

• replace transpose-radix2 with much more general transpose-recurse solver

  Show description for 0129b31
  stevengj

  committed
  0129b31

  Copy full SHA for 0129b31

• rename transpose-inplace to transpose-pairwise, as the algorithm is not restricted to inplace operation

  Show description for b5399f6
  stevengj

  committed
  b5399f6

  Copy full SHA for b5399f6

Commits on Jan 10, 2007

• whoops, some int/INT bugs

  Show description for 1db8349
  stevengj

  committed
  1db8349

  Copy full SHA for 1db8349

• Missing ``static'' keyword.

  Show description for 1f9ce0c
  matteo-frigo

  committed
  1f9ce0c

  Copy full SHA for 1f9ce0c

• Minor cleanup.

  Show description for 5719264
  matteo-frigo

  committed
  5719264

  Copy full SHA for 5719264

Commits on Jan 9, 2007

• fix FAQ Makefile for vpath builds

  Show description for 7c54c72
  stevengj

  committed
  7c54c72

  Copy full SHA for 7c54c72

• interleave twiddle mults with DFTs (should we use dftw?)

  Show description for 31f2376
  stevengj

  committed
  31f2376

  Copy full SHA for 31f2376

Commits on Jan 8, 2007

• simplified (and somewhat sped up) dft-rank1 by exploiting dft-rank1-bigvec

  Show description for 5ee274d
  stevengj

  committed
  5ee274d

  Copy full SHA for 5ee274d

Commits on Jan 7, 2007

• rearranged TRANSPOSED format, numerous speedups

  Show description for 6603c47
  stevengj

  committed
  6603c47

  Copy full SHA for 6603c47

Commits on Jan 5, 2007

• add bench_cost_postprocess to prevent deadlocks in mpi-bench

  Show description for 8efa4e8
  stevengj

  committed
  8efa4e8

  Copy full SHA for 8efa4e8

Commits on Jan 4, 2007

• whoops

  Show description for ad8fbe7
  stevengj

  committed
  ad8fbe7

  Copy full SHA for ad8fbe7

Commits on Jan 3, 2007

• pass proper pointer types as arguments, so that ACX_PTHREAD still works with C++ and -Werror (thanks to Ewald Arnold for the suggestion)

  Show description for f1d13c4
  stevengj

  committed
  f1d13c4

  Copy full SHA for f1d13c4

Commits on Jan 2, 2007

• Renamed [io]vs => [io]vs_by_nbuf, which is more appropriate and would have saved me 30mins debugging.

  Show description for 85662c7
  matteo-frigo

  committed
  85662c7

  Copy full SHA for 85662c7

Commits on Dec 18, 2006

• fixed hc2c for vector-recursion branch

  Show description for 3612cb7
  matteo-frigo

  committed
  3612cb7

  Copy full SHA for 3612cb7

Commits on Dec 5, 2006

• switch to default vector recursion

  Show description for af9505f
  matteo-frigo

  committed
  af9505f

  Copy full SHA for af9505f

Pagination

Previous

Next