This is the mail archive of the gcc-bugs@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

[Bug target/48690] m68k-netbsdelf stack and object alignments are broken


https://gcc.gnu.org/bugzilla/show_bug.cgi?id=48690

Jeffrey A. Law <law at redhat dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P3                          |P5
                 CC|                            |law at redhat dot com
            Summary|gcc-4.3.5 fails for target  |m68k-netbsdelf stack and
                   |m68k                        |object alignments are
                   |                            |broken

--- Comment #36 from Jeffrey A. Law <law at redhat dot com> ---
I'm making this BZ the canonical one for issues with the broken alignments on
the m68k-netbsd port.

One of the problems (identified in BZ 59674) is that if you have fundamental
data alignments that are greater than the stack alignment without dynamic stack
realignment, then things are going to break.  Fundamentally in that situation
there's not a good way for the compiler to ensure that objects on the stack
have the desired alignment because the stack itself may not be properly
aligned.

Changing either the alignment of the objects or the stack alignment
requirements is an ABI change and is thus probably undesirable.

While these issues can be fixed with dynamic stack realignment, nobody from the
GCC team is currently looking at supporting dynamic stack realignment on the
m68k.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]