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

Subversion Repositories or1k

[/] [or1k/] [trunk/] [or1ksim/] [support/] - Rev 1748

Rev

Go to most recent revision | Details | Compare with Previous | Blame

Filtering Options

Clear current filter

Rev Log message Author Age Path
1748 These are all the changes for Or1ksim 0.3.0 release candidate 1. The changes
are explained in the NEWS, README and ChangeLog files. A number of
long-standing bugs are fixed (see the OpenRISC tracker), and the code is
brought up to a consistent standard, following the GNU coding conventions
throughout.

Argument parsing now uses argtable2, and a User Guide has been added.
Documentation throughout has been extended to be compatible with Doxygen,
providing a further level of technical detail on the internals.		 jeremybennett 5871d 14h /or1k/trunk/or1ksim/support
1745 These are the changes to allow or1ksim to build as a library as well as a standalone simulator. The concept of a "generic" peripheral is added, which will commuicate with an external model via upcalls. jeremybennett 5907d 13h /or1k/trunk/or1ksim/support
1744 Changes to bring behavior into line with the current OpenRISC 1000 specification and support GDB 6.8. A couple of small bugs with handling xterms and opening the remote debug channel are also fixed. Header files have been added to sources in Makefile.am files where they are missing, so that "make tags" will include them. Makefile.in files have been regenerated due to these changes. jeremybennett 5908d 13h /or1k/trunk/or1ksim/support
1689 Move do_scheduler() to sched.c. Doesn't result in any performance loss. nogj 6868d 11h /or1k/trunk/or1ksim/support
1688 Don't inline sched_add/sched_remove. This only results in a minimal performance
loss.		 nogj 6868d 11h /or1k/trunk/or1ksim/support
1687 Don't use MMX registers to keep track of the pc and the number of cycles
remaining untill the next job. Keeping the two places that these where used at
concurrent bloated generated code quite a bit and there was no performance
increase resulting from their use.		 nogj 6868d 11h /or1k/trunk/or1ksim/support
1650 Make the loading code not use the debug() function. The coff debug output goes
to the coff channel, the elf ones where pretty useless so I removed them.		 nogj 6868d 12h /or1k/trunk/or1ksim/support
1619 Fixed types in function declaration jcastillo 6903d 23h /or1k/trunk/or1ksim/support
1604 Fix dumphex/dumpverilog to not do unaligned memory access nogj 6930d 13h /or1k/trunk/or1ksim/support
1594 Fix the case of is_power2(0) nogj 6942d 15h /or1k/trunk/or1ksim/support
1589 Make -d channel be equivalent to -d +channel nogj 6949d 00h /or1k/trunk/or1ksim/support
1576 configure updates phoenix 6981d 08h /or1k/trunk/or1ksim/support
1569 Fix segfault when parsing invalid debug options nogj 7003d 15h /or1k/trunk/or1ksim/support
1558 Fix a bug where the scheduler stack would underrun when a job was scheduled run in negative time nogj 7004d 22h /or1k/trunk/or1ksim/support
1557 Fix most warnings issued by gcc4 nogj 7004d 22h /or1k/trunk/or1ksim/support
1555 * Moved log2_int() from cuc/cuc.c as it is usefull for other things aswell.
* Changed code to use log2_int() instead of log2(), which is also a builtin
library function (fixes compile on gcc4).
* Moved is_power2() from sim-config.c to misc.c.		 nogj 7004d 22h /or1k/trunk/or1ksim/support
1545 move sched_next_insn from sim-cmd.c to sched.c. It is also usefull for the pic and the tick timer nogj 7066d 11h /or1k/trunk/or1ksim/support
1541 Print the scheduler jobs when the sched_jobs debug channel has been specified nogj 7066d 11h /or1k/trunk/or1ksim/support
1532 Add pretty spr dumping code nogj 7070d 11h /or1k/trunk/or1ksim/support
1530 Move the checking of the debug channel into the TRACE() macro nogj 7070d 11h /or1k/trunk/or1ksim/support

1 2 Next >

Show All

powered by: WebSVN 2.1.0

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