diff options
| author | charlet <charlet@138bc75d-0d04-0410-961f-82ee72b054a4> | 2010-09-09 12:46:27 +0000 |
|---|---|---|
| committer | charlet <charlet@138bc75d-0d04-0410-961f-82ee72b054a4> | 2010-09-09 12:46:27 +0000 |
| commit | 5e797030173bec0d4f9575a8930c8445e82dae60 (patch) | |
| tree | 95d004c63d3883e1949ae823996d9c6caeea7d2a /gcc/ada/sem_attr.adb | |
| parent | fb3e33e4de2f3d0632c30d29fdc417f25360c671 (diff) | |
| download | gcc-5e797030173bec0d4f9575a8930c8445e82dae60.tar.gz | |
2010-09-09 Robert Dewar <dewar@adacore.com>
* sem_attr.adb: Minor reformatting.
2010-09-09 Thomas Quinot <quinot@adacore.com>
* socket.c (__gnat_socket_ioctl): On Darwin, the req parameter is an
unsigned long, not an int.
2010-09-09 Vincent Celier <celier@adacore.com>
* make.adb, mlib-prj.adb, prj.adb, prj-nmsc.adb, mlib-tgt.adb,
prj-conf.adb, prj-env.adb: Use Display_Name instead of Name whenever
we are not checking for equality of path or file names.
2010-09-09 Ed Schonberg <schonberg@adacore.com>
* exp_util.adb (Remove_Side_Effects): If the expression is a packed
array reference, reset the Analyzed flag so that it is properly
expanded when the resulting object declaration is analyzed.
2010-09-09 Vincent Celier <celier@adacore.com>
* sinput-p.adb (Source_File_Is_Subunit): Return False if X is
No_Source_File.
2010-09-09 Ramon Fernandez <fernandez@adacore.com>
* sysdep.c: The wrSbc8548 BSP in MILS doesn't know anything about the
VX_SPE_TASK option, so disable it.
git-svn-id: svn+ssh://gcc.gnu.org/svn/gcc/trunk@164084 138bc75d-0d04-0410-961f-82ee72b054a4
Diffstat (limited to 'gcc/ada/sem_attr.adb')
| -rw-r--r-- | gcc/ada/sem_attr.adb | 14 |
1 files changed, 9 insertions, 5 deletions
diff --git a/gcc/ada/sem_attr.adb b/gcc/ada/sem_attr.adb index 3eccf93c006..6e7b22ae7b4 100644 --- a/gcc/ada/sem_attr.adb +++ b/gcc/ada/sem_attr.adb @@ -7930,8 +7930,6 @@ package body Sem_Attr is -- didn't permit the access to be declared in the generic -- spec, whereas the revised rule does (as long as it's not -- a formal type). - -- Note that we relax this check in CodePeer mode for - -- compatibility with legacy code. -- There are a couple of subtleties of the test for applying -- the check that are worth noting. First, we only apply it @@ -7951,6 +7949,12 @@ package body Sem_Attr is -- when within an instance, because any violations will have -- been caught by the compilation of the generic unit. + -- Note that we relax this check in CodePeer mode for + -- compatibility with legacy code. + + -- This seems an odd decision??? Why should codepeer mode + -- have a different notion of legality from the compiler??? + elsif Attr_Id = Attribute_Access and then not CodePeer_Mode and then not In_Instance @@ -7969,9 +7973,9 @@ package body Sem_Attr is -- The attribute type's ultimate ancestor must be -- declared within the same generic unit as the -- subprogram is declared. The error message is - -- specialized to say "ancestor" for the case where - -- the access type is not its own ancestor, since - -- saying simply "access type" would be very confusing. + -- specialized to say "ancestor" for the case where the + -- access type is not its own ancestor, since saying + -- simply "access type" would be very confusing. if Enclosing_Generic_Unit (Entity (P)) /= Enclosing_Generic_Unit (Root_Type (Btyp)) |