zhouyuan commented on code in PR #13363:
URL: https://github.com/apache/gluten/pull/13363#discussion_r4323465925
##########
cpp/core/utils/tac/ffor.hpp:
##########
@@ -280,23 +280,57 @@ inline void decodeRt(const uint64_t* in, uint64_t* out,
uint64_t base, size_t n,
detail::kDecodeTable[bw](in, out, base, n);
}
+// AVX2 has no unsigned 64-bit min/max -- vpminuq and vpmaxuq need AVX-512 --
so at the build
+// baseline gcc emulates each comparison with a sign-bit bias, a signed
compare and a variable blend,
+// and the blend lands on the loop-carried accumulator chain. target_clones
has the compiler emit one
+// version of this loop per target, plus an ifunc resolver that picks between
them once at load time:
+// the default clone keeps that emulation, the x86-64-v4 clone collapses it to
one instruction per
+// accumulator. Results are identical either way; this is the same unsigned
min/max.
+//
+// It is its own function because that is what target_clones can clone; where
the attribute expands to
+// nothing gcc is free to inline it back into analyze(), which is what the
previous code did.
+//
+// Multiversioning is on from GCC 12, the first release that can resolve the
level at runtime: GCC 11
+// accepts -march=x86-64-v4 but has no __builtin_cpu_supports for the level,
so it rejects the
+// attribute outright instead of ignoring it. Clang is excluded because the
psABI level names only
+// reach its multiversioning table in LLVM 18 and that combination is untested
here, and the dispatch
+// itself needs an ifunc, hence ELF. Those builds, and non-x86-64 targets,
compile a plain baseline
+// function.
+#ifndef GLUTEN_ENABLE_MULTITARGET_CODE
+#define GLUTEN_ENABLE_MULTITARGET_CODE 1
+#endif
+#if GLUTEN_ENABLE_MULTITARGET_CODE && defined(__x86_64__) && defined(__ELF__)
&& defined(__GNUC__) && \
+ !defined(__clang__) && __GNUC__ >= 12
+#define GLUTEN_FFOR_MULTIVERSION
__attribute__((target_clones("arch=x86-64-v4", "default")))
+#else
+#define GLUTEN_FFOR_MULTIVERSION
+#endif
+
+GLUTEN_FFOR_MULTIVERSION inline void minMaxScan(const uint64_t* data, size_t
n, uint64_t& mn, uint64_t& mx) {
Review Comment:
@hhr293 this seems a valid issue, could you please check?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
