From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from draig.lan ([85.9.250.243]) by smtp.gmail.com with ESMTPSA id 4fb4d7f45d1cf-5dc1861973asm1370688a12.2.2025.01.24.09.15.26 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 24 Jan 2025 09:15:27 -0800 (PST) Received: from draig (localhost [IPv6:::1]) by draig.lan (Postfix) with ESMTP id 1501D5F839; Fri, 24 Jan 2025 17:15:26 +0000 (GMT) From: =?utf-8?Q?Alex_Benn=C3=A9e?= To: Peter Maydell Cc: qemu-arm@nongnu.org, qemu-devel@nongnu.org Subject: Re: [PATCH 02/76] tests/tcg/x86_64/fma: Test some x86 fused-multiply-add cases In-Reply-To: <20250124162836.2332150-3-peter.maydell@linaro.org> (Peter Maydell's message of "Fri, 24 Jan 2025 16:27:22 +0000") References: <20250124162836.2332150-1-peter.maydell@linaro.org> <20250124162836.2332150-3-peter.maydell@linaro.org> User-Agent: mu4e 1.12.8; emacs 29.4 Date: Fri, 24 Jan 2025 17:15:26 +0000 Message-ID: <87r04si15d.fsf@draig.linaro.org> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable X-TUID: fwf05zjc30xC Peter Maydell writes: > Add a test case which tests some corner case behaviour of > fused-multiply-add on x86: > * 0 * Inf + SNaN should raise Invalid > * 0 * Inf + QNaN shouldh not raise Invalid > * tininess should be detected after rounding > > There is also one currently-disabled test case: > * flush-to-zero should be done after rounding > > This is disabled because QEMU's emulation currently does this > incorrectly (and so would fail the test). The test case is kept in > but disabled, as the justification for why the test running harness > has support for testing both with and without FTZ set. > > Signed-off-by: Peter Maydell > --- > tests/tcg/x86_64/fma.c | 109 +++++++++++++++++++++++++++++++ > tests/tcg/x86_64/Makefile.target | 1 + > 2 files changed, 110 insertions(+) > create mode 100644 tests/tcg/x86_64/fma.c > > diff --git a/tests/tcg/x86_64/fma.c b/tests/tcg/x86_64/fma.c > new file mode 100644 > index 00000000000..09c622ebc00 > --- /dev/null > +++ b/tests/tcg/x86_64/fma.c > @@ -0,0 +1,109 @@ > +/* > + * Test some fused multiply add corner cases. > + * > + * SPDX-License-Identifier: GPL-2.0-or-later > + */ > +#include > +#include > +#include > +#include > + > +#define ARRAY_SIZE(x) (sizeof(x) / sizeof((x)[0])) > + > +/* > + * Perform one "n * m + a" operation using the vfmadd insn and return > + * the result; on return *mxcsr_p is set to the bottom 6 bits of MXCSR > + * (the Flag bits). If ftz is true then we set MXCSR.FTZ while doing > + * the operation. > + * We print the operation and its results to stdout. > + */ > +static uint64_t do_fmadd(uint64_t n, uint64_t m, uint64_t a, > + bool ftz, uint32_t *mxcsr_p) > +{ > + uint64_t r; > + uint32_t mxcsr =3D 0; > + uint32_t ftz_bit =3D ftz ? (1 << 15) : 0; > + uint32_t saved_mxcsr =3D 0; > + > + asm volatile("stmxcsr %[saved_mxcsr]\n" > + "stmxcsr %[mxcsr]\n" > + "andl $0xffff7fc0, %[mxcsr]\n" > + "orl %[ftz_bit], %[mxcsr]\n" > + "ldmxcsr %[mxcsr]\n" > + "movq %[a], %%xmm0\n" > + "movq %[m], %%xmm1\n" > + "movq %[n], %%xmm2\n" > + /* xmm0 =3D xmm0 + xmm2 * xmm1 */ > + "vfmadd231sd %%xmm1, %%xmm2, %%xmm0\n" > + "movq %%xmm0, %[r]\n" > + "stmxcsr %[mxcsr]\n" > + "ldmxcsr %[saved_mxcsr]\n" > + : [r] "=3Dr" (r), [mxcsr] "=3Dm" (mxcsr), > + [saved_mxcsr] "=3Dm" (saved_mxcsr) > + : [n] "r" (n), [m] "r" (m), [a] "r" (a), > + [ftz_bit] "r" (ftz_bit) > + : "xmm0", "xmm1", "xmm2"); > + *mxcsr_p =3D mxcsr & 0x3f; > + printf("vfmadd132sd 0x%" PRIx64 " 0x%" PRIx64 " 0x%" PRIx64 > + " =3D 0x%" PRIx64 " MXCSR flags 0x%" PRIx32 "\n", > + n, m, a, r, *mxcsr_p); > + return r; > +} > + > +typedef struct testdata { > + /* Input n, m, a */ > + uint64_t n; > + uint64_t m; > + uint64_t a; > + bool ftz; > + /* Expected result */ > + uint64_t expected_r; > + /* Expected low 6 bits of MXCSR (the Flag bits) */ > + uint32_t expected_mxcsr; > +} testdata; > + > +static testdata tests[] =3D { > + { 0, 0x7ff0000000000000, 0x7ff000000000aaaa, false, /* 0 * Inf + SNa= N */ > + 0x7ff800000000aaaa, 1 }, /* Should be QNaN and does raise Invalid = */ > + { 0, 0x7ff0000000000000, 0x7ff800000000aaaa, false, /* 0 * Inf + QNa= N */ > + 0x7ff800000000aaaa, 0 }, /* Should be QNaN and does *not* raise In= valid */ > + /* > + * These inputs give a result which is tiny before rounding but which > + * becomes non-tiny after rounding. x86 is a "detect tininess after > + * rounding" architecture, so it should give a non-denormal result a= nd > + * not set the Underflow flag (only the Precision flag for an inexact > + * result). > + */ > + { 0x3fdfffffffffffff, 0x001fffffffffffff, 0x801fffffffffffff, false, > + 0x8010000000000000, 0x20 }, > + /* > + * Flushing of denormal outputs to zero should also happen after > + * rounding, so setting FTZ should not affect the result or the flag= s. > + * QEMU currently does not emulate this correctly because we do the > + * flush-to-zero check before rounding, so we incorrectly produce a > + * zero result and set Underflow as well as Precision. > + */ > +#ifdef ENABLE_FAILING_TESTS > + { 0x3fdfffffffffffff, 0x001fffffffffffff, 0x801fffffffffffff, true, > + 0x8010000000000000, 0x20 }, /* Enabling FTZ shouldn't change flags= */ > +#endif We could extend the multiarch/float_madds test to handle doubles as well (or create a new multiarch test). If the right FMA instructions can be forced via cflags you could then add a specific binary to exercise the vfmadd231sd instruction like we do for neon: float_madds: CFLAGS+=3D-mfpu=3Dneon-vfpv4 The test does basic testing but ideally you would add a reference output to check against. > +}; > + > +int main(void) > +{ > + bool passed =3D true; > + for (int i =3D 0; i < ARRAY_SIZE(tests); i++) { > + uint32_t mxcsr; > + uint64_t r =3D do_fmadd(tests[i].n, tests[i].m, tests[i].a, > + tests[i].ftz, &mxcsr); > + if (r !=3D tests[i].expected_r) { > + printf("expected result 0x%" PRIx64 "\n", tests[i].expected_= r); > + passed =3D false; > + } > + if (mxcsr !=3D tests[i].expected_mxcsr) { > + printf("expected MXCSR flags 0x%x\n", tests[i].expected_mxcs= r); > + passed =3D false; > + } > + } > + return passed ? 0 : 1; > +} > diff --git a/tests/tcg/x86_64/Makefile.target b/tests/tcg/x86_64/Makefile= .target > index d6dff559c7d..be20fc64e88 100644 > --- a/tests/tcg/x86_64/Makefile.target > +++ b/tests/tcg/x86_64/Makefile.target > @@ -18,6 +18,7 @@ X86_64_TESTS +=3D adox > X86_64_TESTS +=3D test-1648 > X86_64_TESTS +=3D test-2175 > X86_64_TESTS +=3D cross-modifying-code > +X86_64_TESTS +=3D fma > TESTS=3D$(MULTIARCH_TESTS) $(X86_64_TESTS) test-x86_64 > else > TESTS=3D$(MULTIARCH_TESTS) --=20 Alex Benn=C3=A9e Virtualisation Tech Lead @ Linaro