-
Notifications
You must be signed in to change notification settings - Fork 23
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Parallel rns fgemv #273
base: master
Are you sure you want to change the base?
Parallel rns fgemv #273
Conversation
…on into pfgemv which will no more require to be labeled with PAR_BLOCK
…gemv-mp has been restructured for different parameter values
…r benchmarking fgemv in the rns field
…ervable speedup for rns fgemv
@ClementPernet Looks like all Travis builds are failing even on master. Have there been any updates in the configuration recently? |
@@ -158,7 +158,7 @@ AC_PROG_LIBTOOL | |||
AC_PROG_EGREP | |||
AC_PROG_SED | |||
# newer libtool... | |||
LT_PREREQ([2.4.3]) | |||
LT_PREREQ([2.4.2]) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
You should not have to change this
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I had to do so because I once got error with travis as if the libtool version was not corret. By changing the version to a lower number, I passed the travis compilation without the strange build failure.
{ | ||
for (size_t j=0; j<N; ++j) { | ||
const Givaro::Integer & x(A[i*lda+j]); | ||
if (Givaro::absCompare(x,vmax[i])>0){ vmax[i] = x;} |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Updating vmax[i] in each parallel task creates contention on every operation. This code is likely not parallel at all.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yes not so correctly parallelized but timing showed a few speedup. I cannot see any easier way to find the local max value for each thread then search for the global max value outside the parallel region.
@@ -410,7 +410,9 @@ namespace FFLAS { | |||
else if (!std::is_same<Field,Givaro::ModularBalanced<float> >::value){ | |||
if (F.characteristic() < DOUBLE_TO_FLOAT_CROSSOVER) | |||
return Protected::fgemm_convert<Givaro::ModularBalanced<float>,Field>(F,ta,tb,m,n,k,alpha,A,lda,B,ldb,beta,C,ldc,H); | |||
else if (!std::is_same<Field,Givaro::ModularBalanced<double> >::value && 16*F.cardinality() < Givaro::ModularBalanced<double>::maxCardinality()) | |||
else if (!std::is_same<Field,Givaro::ModularBalanced<double> >::value && | |||
!std::is_same<Field,Givaro::ModularBalanced<double> >::value && |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
why is this line duplicated?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
If I remember well, we did this together to correct some compile time error, but I could not remember of the exact reason.
Implementation of fgemv for rns with corresponding helpers