Closed mborland closed 1 month ago
Attention: Patch coverage is 95.54455%
with 9 lines
in your changes missing coverage. Please review.
Project coverage is 94.06%. Comparing base (
66b362a
) to head (9ed738f
). Report is 3 commits behind head on develop.
Files with missing lines | Patch % | Lines |
---|---|---|
include/boost/math/special_functions/gamma.hpp | 90.62% | 6 Missing :warning: |
include/boost/math/tools/toms748_solve.hpp | 89.65% | 3 Missing :warning: |
This begins to remove earlier workarounds for NVRTC now that support is more robust. In this PR workarounds were removed from gamma so now instead of having just
tgamma
, andlgamma
on device we have the whole gamma family including:tgamma_ratio
,tgamma_delta_ratio
,gamma_p_inv
,gamma_p_inva
,gamma_q_inv
,gamma_q_inva
,gamma_p_derivative
,igamma
,igamma_inv
,igamma_inva
. This additional support allows us to add GPU support to the chi squared distribution and weibull distributions. We also gain root finding support with toms748 now being device capable.This PR is a bit uglier than other because NVRTC does not really support forward declarations.
gamma.hpp
used these extensively so we either call the CUDA standard library where we can, or apply the correct pre-processing to call detail functions. A little bit of recursion was eliminated at the cost of additional dispatching functions.Completed device CI runs can be found here: https://github.com/cppalliance/cuda-math/pull/18.
CC: @steppi, @dschmitz89, @izaid.