OpenCores
URL https://opencores.org/ocsvn/or1k/or1k/trunk

Subversion Repositories or1k

[/] [or1k/] [trunk/] [gdb-5.3/] [sim/] [ppc/] [BUGS] - Rev 1765

Compare with Previous | Blame | View Log

ChangeLog

See the ChangeLog file looking for lines taged with the word FIXME.


COREFILE.C:

The implementation of corefile.c (defined by corefile.h) isn't the
best.  It is intended to be functionaly correct rather than fast.  One
option being considered is to add a data cache to reduce the overhead
of the most common case of data read/writes.


VEA:

Missing VEA system calls.

ppc-instructions:

Missing or commented out instructions.


64bit:

64bit target untested. 64bit host broken.  For instance use of scanf
"%x", &long long.



hw_*.c:

Better and more devices.

PORTABILITY:

(Notes taken from Michael Meissner): Heavy use of the ## operator -
fix using the clasic X/**/Y hack; Use of the signed keyword.  In
particular, signed char has no analogue in classic C (though most
implementations of classic C use signed chars); Use of long long which
restricts the target compiler to be GCC.


TRACING:

debug.c: Macro's should be extended to include:

        IS_*TRACE:      True if tracing enabled
        *TRACE_PREFIX:  Outputs just the prefix line

hw_trace.c: Flush, replace with a psim_set_tracing or some
        such program.


CIA/NIA:

Replace with functions to return/increment the CIA?



SMP & GDB:

GDB doesn't understand SMP!




OVERALL STRUCTURE:

A new file pstruct.h is to be created that contains a single flat data
structure containing:

                pstruct {
                        events;
                        core;
                        processor[nr_cpus];
                        monitor;
                        devices;
                        trace;
                }

The CPU's structure, in turn would contain the VM sub structures.

When SMP==0, everything would have PSTRUCT passed.  In SMP mode,
however, there are two choices: PSTRUCT + CPU_NR or PROCESSOR.  I
suspect the latter is better.

It is believed that this would significantly improve performance (at
the price of reduced control over object scope).




IGEN:

Igen at present can't do the following:

        o       duplication is an all or nothing afair.

                It should be configurable according to
                the instruction or the sub-table.


        o       Due to the naming, only a single generated
                simulator can be included in a program.

                IGEN should be able to generate multiple
                engines that can all be included in a program

        o       handle alternate architectures.


        o       Igen should support the generation of a
                disasembler and posibly an assembler.

                I suggest that the table be extended to
                include, for each instruction, additional
                lines describing the extual format of the
                instruction.

                One possible format is:

                        "mtlr %RS":SPR.something
                        "mtspr %SPR, %RS"

Compare with Previous | Blame | View Log

powered by: WebSVN 2.1.0

© copyright 1999-2024 OpenCores.org, equivalent to Oliscience, all rights reserved. OpenCores®, registered trademark.