| RE: [cdt-dev] [Bug?] Signed registers in registers view (nasty in hex) |
|
The bug seems to be in
SourceManager.java (package org.eclipse.cdt.debug.mi.core.cdi). In the method toCDIType, we return
the type as “Long” for 32-bit values. Shouldn’t this be
“Int” ? <snip>
// GDB has some special types for int
if (typename.equals("int8_t")) { //$NON-NLS-1$
return new CharType(target, typename);
} else if (typename.equals("int16_t")) { //$NON-NLS-1$
return new ShortType(target, typename);
} else if (typename.equals("int32_t")) { //$NON-NLS-1$
return new IntType(target,
typename); // works
//return new LongType(target,
typename); // original code - doesn’t
work
} else if (typename.equals("int64_t")) { //$NON-NLS-1$
return new LongLongType(target, typename);
} else if (typename.equals("int128_t")) { //$NON-NLS-1$
return new IntType(target, typename); // ????
} </snip> Thanks, Sridhar --- Sridhar Sailappan, Senior Software Engineer, Xilinx Inc, -----Original Message----- Hi, gdb has so-called "natural" format which is the default
format for the Registers view. This means that signed is the "natural"
format for registers in gdb. Conversion to hex is done by CDT and is based on the register's type
and endiannes, so this can be a bug in CDT. Thanks, Mikhail Khodjaiants ----- Original Message ----- From: "Sascha Radike" <sradike@xxxxxxxxxxxx> To: <cdt-dev@xxxxxxxxxxx> Sent: Thursday, July 06, 2006 3:22 PM Subject: [cdt-dev] [Bug?] Signed registers in registers view (nasty in
hex) > Hi, > > I just had a look at the registers view (Eclipse RC7 + CDT 3.2
HEAD two > weeks ago or so, GDB 6.4) and noticed that registers show up as
signed (-1 > instead of 4294967295). > > I don't think this is a popular view for CPU registers ? > > Anyway, it gets ugly when I use the hex view as it shows up as
long! > > Example: > -4096 shows up as 0xfffffffffffff000 > > The registers are only 32 bit. > > When I use the "Cast to type" to cast to "unsigned
int" it shows up > correctly. But shouldn't the registers show up correctly by
default? > > CDT bug or GDB (once again)? Or am I just blind ? > > Thanks > > Sascha > > _______________________________________________ > cdt-dev mailing list > cdt-dev@xxxxxxxxxxx > https://dev.eclipse.org/mailman/listinfo/cdt-dev _______________________________________________ cdt-dev mailing list cdt-dev@xxxxxxxxxxx https://dev.eclipse.org/mailman/listinfo/cdt-dev |