Page MenuHomeFreeBSD

Use C99 types in MSDOSFS system code in anticipation of makefs msdos support
ClosedPublic

Authored by guest-svmhdvn on May 19 2017, 3:51 PM.
Tags
None
Referenced Files
Unknown Object (File)
Oct 14 2024, 9:13 PM
Unknown Object (File)
Oct 14 2024, 8:37 PM
Unknown Object (File)
Oct 14 2024, 8:36 PM
Unknown Object (File)
Oct 14 2024, 8:08 PM
Unknown Object (File)
Sep 30 2024, 5:32 PM
Unknown Object (File)
Sep 30 2024, 8:53 AM
Unknown Object (File)
Sep 20 2024, 7:25 PM
Unknown Object (File)
Sep 12 2024, 7:17 PM
Subscribers

Details

Summary

General types cleanup in msdosfs code only. This also helps reduce diffs with NetBSD msdosfs libraries and msdos support in makefs.

Diff Detail

Lint
Lint Skipped
Unit
Tests Skipped

Event Timeline

sys/fs/msdosfs/denode.h
107–108

I noticed NetBSD still has u_long here (and elsewhere in this file)

sys/fs/msdosfs/denode.h
107–108

Do we want to prioritize the diff reduction or consistency with the rest of our code and style(9) ("The project is slowly moving to use the ISO/IEC 9899:1999 (``ISO C99'') unsigned integer identifiers of the form uintXX_t...")? It seems they are sometimes using uXXX_t and other times using u_XXX_t.

sys/fs/msdosfs/denode.h
107–108

We definitely want to use C99 fixed width types where the exist -- so we'll prefer int32_t to u_int32_t always.

This doesn't apply to the other convenience typedefs though -- u_long vs unsigned long. Does NetBSD's makefs msdos support include the appropriate header(s) to define u_long already?

sys/fs/msdosfs/denode.h
107–108

Ah my mistake, this is simply a convenience typedef, NetBSD has the same typedefs set up in sys/types.h. I will revert these changes for the u_long and u_char typedefs.

  • Removed whitespace and style changes on lines other than the ones affected in this diff
  • Kept u_long, u_char, etc. as is
This revision is now accepted and ready to land.May 19 2017, 5:39 PM
This revision was automatically updated to reflect the committed changes.