----------------------------------------------------------------- Changes in release 12 ----------------------------------------------------------------- - reworked GCC-toolchain: Valid mips*-elf-binaries are now compiled and linked using a single mips*-elf-gcc command. To use: Copy $THIS_REALEASE/bsp/toolchain/specs to $MIPS_GCC_PREFIX/lib/gcc/mipsel-elf/$GCC_VERSION/, and see bsp/examples/Makefile gcc usage GCC-provided crt*-functions are now called without problems. In summary: The toolchain is now more compliant to the intentional use of GCC. - mips_biu.vhd: Addresses 0xA0000000..0xBFFFFFFF are now un-dcached for memory-mapped I/O, which is now compliant to literature. - gpio_wb.vhd: Timer reset is now removed. This helps to keep current time in case of CPU reset. - new bsp-example source: gunzip - updated drawing 'mips_jbus_timing.pdf' - added drawing 'memory_map.pdf' - moved COP0 from mips_pipeline.vhd to mips_top.vhd. Created new cop-interface for later COPz connection. - added I/D-cache info (size and linesize) to upper 16 bits of PrID-Register. Example of reading cache info is done in bsp/examples/testbench::PrintCPUinfo() - simplified instruction decoder in mips_instr.vhd. Now there are less warnings during synthesis. Removed idecode_rom.vhd from project as consequence. - added cop0 instructions to invalidate I-cache and D-Cache - added copz instructions lwcz and swcz - gpio_wb.vhd: added 32 bit timer with interrupt and auto reload - added debugger in libsys. New files are mipsdis.c/h and dbg.c/h - exception handlers can now be registered to any exception type - bugfix: cop0 registers were not written during exception commit. Cop registers now behave like general registers. - added pin 'eb' to mips_top.vhd to select CPU-endianess at reset time (eb = 0 : little-endian, eb = 1 : big-endian) - Status register bit 'RE'= 1 (bit 25) reverses endianess in user-mode - Status register bit 'TS' (TLB Shutdown, bit 21) is tied high to indicate the absence of a MMU. - uart_wb.vhd: added interrupt enable - Interrupt pending flags (IP) are not anymore masked by Interrupt mask flags (IM) in H/W. Masking has to be done in S/W. H/W masking was removed because it isn't MIPS-compliant. - added NMI pin (more MIPS compliant). Asserting causes exception. Exception vector is reset (0xBFC00000). Using NMI as reset is dangerous because periphery is not reseted. Unitialized/unreseted periphery may cause trouble (e.g. firing unwanted interrupts) - Bootloader: Startup now invalidates I/D-Caches at boot - Bugfix: register bypass with same target register, used for unaligned loads (LWL, LWR), was incorrect. Newlib's memcpy() (MIPS-optimized) uses this. Consectutive load instructions without nops (which is legal) using same target register like lw $1, 0(mem) lwl $1, 1(mem) or lwl $1, 0(mem) lwr $1, 1(mem) or similar, lead to incorrect result. Now it'S fixed. ----------------------------------------------------------------- Changes in release 11 ----------------------------------------------------------------- - bug-fix in mips_pipeline: pc_is_branch is now initialized at reset (fixes CPU unrecoverable state after reset) - reverted "optimization" in mips_dcache, which behaved buggy during exceptions - new example source: test_exception - minor changes in libsys - added async_port_timing in doc - examples/testbench: date set now really works ----------------------------------------------------------------- Changes in release 10b ----------------------------------------------------------------- - fixed compiler bug (missing nop after lw) in bootloader, which caused exception during boot ----------------------------------------------------------------- Changes in release 10a ----------------------------------------------------------------- async_port_wb: - added byte enable for async_port bootloader.c - fixed compiler bug (missing nop after lw) mips-core - SEL_O lines are also asserted correctly during reads tools/romgen - TCL-Jtage now compatible with Chipscope 10.1 mips_sys.c - use byte enable for sram and flash