For example gcc’s libgfortran is missing a libquadmath build dependency. It is natural not to encounter it in real world as libquadmath is usually built along with other small runtimes way before g++ or gfortran is ready.
他的基本想法是把 target 的順序打亂掉,也就是在有指定 --shuffle 時,不一定會照 a -> b -> c 的順序往下遞迴,而可能會是 c -> b -> a 或是其他的順序:
all: a b c
這樣對於抓那些在 -j 平行編譯時會出包的套件也很有幫助,不需要在 -j 開很大的情況下才能重製問題,而是平常就有機會在 CI 環境下被抓出來。
I figured that since we were to replace that matrix function anyway, I could try replacing it with XMMatrixInverse being a “modern” replacement for D3DXMatrixInverse. XMMatrixInverse also uses SSE2 instructions so it should be equally optimal to the D3DX function, but I was nearly sure it would break the same way.
Here’s Intel versus AMD relative error of RCPPS instruction: http://const.me/tmp/vrcpps-errors-chart.png AMD is Ryzen 5 3600, Intel is Core i3 6157U.
Over the complete range of floats, AMD is more precise on average, 0.000078 versus 0.000095 relative error. However, Intel has 0.000300 maximum relative error, AMD 0.000315.
Both are well within the spec. The documentation says “maximum relative error for this approximation is less than 1.5*2^-12”, in human language that would be 3.6621E-4.
Source code that compares them by creating 16GB binary files with the complete range of floats: https://gist.github.com/Const-me/a6d36f70a3a77de00c61cf4f6c17c7ac
至於為什麼會生出 NaN 的原因,沒找出來還是有點可惜,不過這個解法還行,就是「新版的 library 既然沒問題,就大家也不要太計較舊版的問題」的概念...