That's scheduled for GCC 13 and was found by Sandra and Frederik,

'omp scan' has undergone quite some transformation:

In 5.0 it was added with a preceding and succeeding structured block.

In 5.1, 'structured block' was replaced by 'structured-block-sequence'
defined as "...a sequence of two or more executable statements..."

In 5.2, the s.b.s. became "... a sequence of zero or more executable
statements ...". While there are restrictions against orphaned
separating directives, having zero statements seems to be fine (albeit odd).

The C/C++ parser already permitted >= executable statements (albeit the
testcases didn't use those – and used {...} to enclose multiple
executable statements). The Fortran parser was very strict and did
require exactly one single executable statement before/after the scan. —
Now both accept >=0.

I have opt for adding a warning if there are zero statements before or
after the scan.

And [C/C++], while technically 'for (...)' / "#pragma omp scan ..." is
valid ("for" is followed by a structured block), the before and with the
attached patch there is an error if the final-loop-body does not have
enclosing curly braces; I don't feel like changing the parser for such
an odd use – and for any sensible use, the curly braces are required.

Thoughts? Comments?

* * *

Warning: It seems as if nearly all warnings related to OpenMP are "0"
and on by default with very few exceptions. I am wondering whether we
should add a -Wopenmp and change all those 0 to warn_openmp to make it
easier for the user to -Wno-openmp or -Werror=openmp - or just to '2>&1
|grep Wopenmp' to find them.

Thoughts?

* * *

Tobias
-----------------
Siemens Electronic Design Automation GmbH; Anschrift: Arnulfstraße 201, 80634 München; Gesellschaft mit beschränkter Haftung; Geschäftsführer: Thomas Heurung, Frank Thürauf; Sitz der Gesellschaft: München; Registergericht München, HRB 106955