<html> <head> <base href="https://bugs.freedesktop.org/"> </head> <body> <div> <a class="bz_bug_link bz_status_NEW " title="NEW - fp64 division gives imprecise results" href="https://bugs.freedesktop.org/show_bug.cgi?id=110435#c3">Comment # 3</a> on <a class="bz_bug_link bz_status_NEW " title="NEW - fp64 division gives imprecise results" href="https://bugs.freedesktop.org/show_bug.cgi?id=110435">bug 110435</a> from <a class="email" href="mailto:b7.10110111@gmail.com" title="Ruslan Kabatsayev <b7.10110111@gmail.com>"> Ruslan Kabatsayev</a> <pre>The problem appears to be with lower_rcp from nir_lower_double_ops.c. Namely, it's supposed to use two Newton-Raphson steps to improve precision of initial single-precision rcp. But instead of actually doing the steps, it employs FMA operation, with a mistake. Namely, the correct expression should be (as noted in the comment, thanks to it being present): x_new = x + x * (1 - x*src). But actual implementation is ra = nir_ffma(b, ra, nir_ffma(b, ra, src, nir_imm_double(b, -1)), ra), which is equivalent to x_new = x - x * (1 - x*src). Notice the minus sign before the outermost multiplication. I'm not sure how to change sign of `ra` to check whether this would completely fix the problem, but at least in an equivalent-code test in C++ I've reproduced the problem, and the change ra=std::fma(ra, std::fma(ra, src, -1.), ra); to ra=std::fma(-ra, std::fma(ra, src, -1.), ra); made the difference between 1./src and the rcp-via-float in the C++ test go away (for src=5.44786569377455 I used in the test).</pre> </div> <hr> You are receiving this mail because: <ul> <li>You are the QA Contact for the bug.</li> <li>You are the assignee for the bug.</li> </ul> </body> </html>