Hi! Thanks for the review! Fixed your nits, the branch is force pushed. -- Best regards, Maxim Kokryashkin     >  >>Max, >> >>Thanks for the patch! LGTM, but there are some nits below. >> >>On 03.02.23, Maksim Kokryashkin via Tarantool-patches wrote: >>> Follows up commit 9b6c0cd8eafdd2e5a8a7ac4b33f6e33b3d8a93b9 >>> ("Don't compile math.modf() anymore."). >>> >>> This patch introduces the test case for the issue with >>> compiled `math.modf`, that was fixed by disabling the >>> compilation for it. >>> >>> Part of tarantool/tarantool#7230 >> >>Minor: It follows up the ticket, not a part of it. >> >>> --- >>> Branch: https://github.com/tarantool/luajit/tree/fckxorg/modf-test >>> PR: https://github.com/tarantool/tarantool/pull/8251 >>> >>> test/tarantool-tests/math-modf.test.lua | 31 +++++++++++++++++++++++++ >>> 1 file changed, 31 insertions(+) >>> create mode 100644 test/tarantool-tests/math-modf.test.lua >>> >>> diff --git a/test/tarantool-tests/math-modf.test.lua b/test/tarantool-tests/math-modf.test.lua >>> new file mode 100644 >>> index 00000000..2aaa5189 >>> --- /dev/null >>> +++ b/test/tarantool-tests/math-modf.test.lua >>> @@ -0,0 +1,31 @@ >>> +local tap = require('tap') >>> +local test = tap.test('math-modf') >>> +test:plan(1) >>> + >>> +local function isnan(x) >>> + return x ~= x >>> +end >>> + >>> +local function array_is_consistent(res) >>> + for i = 1, #res - 1 do >>> + if res[i] ~= res[i + 1] and not (isnan(res[i]) and isnan(res[i + 1])) then >>> + return false >>> + end >>> + end >>> + return true >>> +end >>> + >>> +jit.opt.start('hotloop=1') >>> +jit.on() >> >>This line is excess. >> >>> + >>> +local modf = math.modf >>> +local inf = math.huge >>> + >>> +local r1 = {nil, nil, nil, nil} >>> +local r2 = {nil, nil, nil, nil} >> >>Minor: I'd rather use table.new here to make it clearer. >> >>> + >>> +for i = 1, 4 do >>> + r1[i], r2[i] = modf(inf) >>> +end >>> + >>> +test:ok(array_is_consistent(r1) and array_is_consistent(r2), 'wrong modf') >>> -- >>> 2.37.1 (Apple Git-137.1) >>> >> >>-- >>Best regards, >>IM >