Lines Matching refs:shadow

75 ; load followed by cmp: check that we load the shadow and call __msan_warning_with_origin.
103 ; Check that we store the shadow for the retval.
113 ; Check that we get the shadow for the retval.
128 ; Check that we generate PHIs for shadow.
155 ; Compute shadow for "x << 10"
173 ; Compute shadow for "10 << x": it should have 'sext i1'.
289 ; Check that we propagate shadow for "select"
336 ; arguments. Select condition shadow is sign-extended to the vector type and
337 ; mixed into the result shadow.
417 ; (2nd arg shadow is checked, 1st arg shadow is propagated)
433 ; Check that fdiv, unlike udiv, simply propagates shadow.
449 ; Check that fneg simply propagates shadow.
465 ; Check that we propagate shadow for x<0, x>=0, etc (i.e. sign bit tests)
516 ; Check that we propagate shadow for x<=-1, x>-1, etc (i.e. sign bit tests)
567 ; Check that we propagate shadow for x<0, x>=0, etc (i.e. sign bit tests)
582 ; Check that we propagate shadow for x<=-1, x>0, etc (i.e. sign bit tests)
598 ; Check that we propagate shadow for unsigned relational comparisons with
617 ; Check that loads of shadow have the same alignment as the original loads.
645 ; Check that the same bit manipulation is applied to the shadow values.
646 ; Check that there is a zero test of the shadow of %idx argument, where present.
700 ; Check that shadow of such vector is a vector of integers.
850 ; argument shadow is a compile-time zero constant (which is always the case
867 ; Test argument shadow alignment
890 ; First element shadow
897 ; Second element shadow
907 ; Test shadow propagation for aggregates passed through ellipsis.
975 ; shadow before the call, we don't need to anymore.
989 ; For "musttail" calls we can not insert any shadow manipulating code between
1006 ; For "musttail" calls we can not insert any shadow manipulating code between