Thanks for the pointer! I think Eric was a little bit too generous last time and formatted the commit log for me =P.
I have added the appropriate entries to the Changelog's, and have included those in the patch (attached), which is also updated to master/HEAD, hopefully that is the most helpful approach. Cheers, Richard > -----Original Message----- > From: Arnaud Charlet <char...@adacore.com> > Sent: March 9, 2021 10:51 AM > To: Richard Wai <rich...@annexi-strayline.com> > Cc: gcc-patches@gcc.gnu.org; 'Bob Duff' <d...@adacore.com>; Arnaud > Charlet <char...@adacore.com> > Subject: Re: [PATCH] Ada: hashed container Cursor type predefined equality > non-conformance > > > Oops, I wrote "effected" when I meant "affected". Of course, despite > > reading it several times before sending, I only noticed it after sending. > > > > Edited version: > > > > --- > > > > * libgnat/a-cohase.ads, libgnat/a-cohase.adb, libgnat/a-cohama.ads, > > libgnat/a-cohama.adb, libgnat/a-cihama.ads, libgnat/a-cihama.adb: > > Ensure that Cursor objects always have their "Position" component set > > to ensure predefined equality works as required; synchronize comments > > for the Cursor type defintion across affected packages, and add > > comments to explain the importance of updating the Position component. > > * testsuite/gnat.dg/containers2.adb: New regression test to ensure > > that all hashed containers obey predefined equality requirements for > Cursor objects. > > OK. The commit log follows strict rules, and in particular you need to list the > functions (And types modified), e.g: > > * libgnat/a-cohase.ads, libgnat/a-cohase.adb, libgnat/a-cohama.ads, > libgnat/a-cohama.adb, libgnat/a-cihama.ads, libgnat/a-cihama.adb > (Cursor, > Insert, Find): ... > > See https://gcc.gnu.org/codingconventions.html#ChangeLogs for more > details. > > OK with the above commit log amended. > > Arno
container_cursor_equality_20210309.patch
Description: Binary data
