Commit message (Collapse) | Author | Age | ||
---|---|---|---|---|
... | ||||
* | Simplify how we define exceptions | Karel Kočí | 2018-01-01 | |
| | | | | | Exceptions definition was pretty much just copy and past. That can do preprocessor for us too so use it for it. | |||
* | Update TODO | Karel Kočí | 2018-01-01 | |
| | ||||
* | Tiny typo | Karel Kočí | 2018-01-01 | |
| | ||||
* | Add cpu_trap test | Karel Kočí | 2018-01-01 | |
| | ||||
* | cli: extend tracer and implement reporter | Karel Kočí | 2018-01-01 | |
| | ||||
* | Fix how we pass exception | Karel Kočí | 2018-01-01 | |
| | | | | | | Previous implementation reseted caught exception to type QtMipsException so we lost possibility to check for object type later on. | |||
* | Build.sh should fail if one of commands fail | Karel Kočí | 2018-01-01 | |
| | ||||
* | Remove some obsolete files | Karel Kočí | 2018-01-01 | |
| | ||||
* | Rework compilation so we don't have to use LD_LIBRARY_PATH | Karel Kočí | 2018-01-01 | |
| | ||||
* | Add ability to specify system mips toolchain in tests | Karel Kočí | 2017-12-30 | |
| | ||||
* | Drop jump predictions | Karel Kočí | 2017-12-30 | |
| | | | | | | | | Although there are branch likely instruction the whole mips instruction set is designed around static jump predictions. The usage of delay slot ensures that we don't have to handle jump predictions on cpu it self at all. Adding it on top of that makes little to no sense although that was the original plan. | |||
* | Add missing programmemory.ui | Karel Kočí | 2017-12-21 | |
| | ||||
* | A lot of small changes in gui | Karel Kočí | 2017-12-21 | |
| | | | | | | Added machine status. Widgets now hold size and registers dock has now scrollbars. And more... | |||
* | Some pleanups and small fixes | Karel Kočí | 2017-12-21 | |
| | | | | | These are just code fixes. Shouldn't change anything but makes code cleaner. | |||
* | Update dependencies | Karel Kočí | 2017-12-21 | |
| | | | | | Yep it now works only with Qt5, no deep idea why but I probably used somewhere something that is not compatible. | |||
* | Implement alu for coreview | Karel Kočí | 2017-12-17 | |
| | ||||
* | Prepend include guard of coreview header files with COREVIEW_ | Karel Kočí | 2017-12-17 | |
| | ||||
* | Use drawPolygon more and better way | Karel Kočí | 2017-12-17 | |
| | ||||
* | Use global includes in qtmips_machine | Karel Kočí | 2017-12-17 | |
| | ||||
* | Put qtmips_machine to machine namespace | Karel Kočí | 2017-12-17 | |
| | ||||
* | Revert "Suppress some warning" | Karel Kočí | 2017-12-17 | |
| | | | | This reverts commit 5bf637a429bbcf09092b8d189010c50d61c3562f. | |||
* | Drop removed file from pro file | Karel Kočí | 2017-12-16 | |
| | ||||
* | Implement some barebone registers dock | Karel Kočí | 2017-12-15 | |
| | ||||
* | Implement few initial graphic elements | Karel Kočí | 2017-12-15 | |
| | ||||
* | Change default tab in new dialog | Karel Kočí | 2017-12-15 | |
| | ||||
* | Fix test for JR | Karel Kočí | 2017-12-12 | |
| | ||||
* | Implement some store and load instructions | Karel Kočí | 2017-12-12 | |
| | ||||
* | Add missing exception definition | Karel Kočí | 2017-12-12 | |
| | ||||
* | Fix immediate alu operation | Karel Kočí | 2017-12-12 | |
| | | | | | There should be a sign extension to 32bit when doing immediate operations. | |||
* | Tiny cleanup | Karel Kočí | 2017-12-12 | |
| | ||||
* | Add AC_NONE to MemoryAccess::AccessControl | Karel Kočí | 2017-12-12 | |
| | ||||
* | Add memory type abstract function | Karel Kočí | 2017-12-12 | |
| | | | | | | This functions can be used to write or read 32bit values from memory and memory on its own does signextends for example depending on passed control value. | |||
* | Implement branch and jump instructions | Karel Kočí | 2017-12-12 | |
| | ||||
* | Don't create memory section on read_byte | Karel Kočí | 2017-12-12 | |
| | ||||
* | Acknowledge that we are not doing exact instruction match | Karel Kočí | 2017-12-12 | |
| | ||||
* | Suppress some warning | Karel Kočí | 2017-12-12 | |
| | | | | | You can call it a fix but in reality it has same effect but I accept that this is better. | |||
* | Fix qt5 qmake not found on some distributions and todo | Karel Kočí | 2017-12-11 | |
| | ||||
* | Set MIPS_PREFIX to contain tailing dash | Karel Kočí | 2017-12-11 | |
| | | | | | | | It's common practice to have cross compilation prefix with dash as it allows you to not use any prefix at all and still have valid gcc call (although we are checking if we have it set at all but that doesn't matter for now). | |||
* | Another but same fix for cache | Karel Kočí | 2017-12-11 | |
| | | | | Segfault happening had same cause as previous problem so fix is similar. | |||
* | Fix problem with no such file or directory | Karel Kočí | 2017-12-11 | |
| | | | | | Yep when object was instantiated it wasn't copying elf_path so now fixed (should also not forget to do the same with other options too). | |||
* | Change used gcc version because of compilation error in ubsan.c | Karel Kočí | 2017-12-11 | |
| | | | | | | | | Compiling gcc 6.3.0 with gcc 7.2.0 results to error: gcc-6.3.0/gcc/ubsan.c:1474:23: error: ISO C++ forbids comparison between pointer and integer So I changed used version to 5.4.0 because of that. Probably in future I will update ct-ng and use gcc 7.2.0 but not now. | |||
* | Add crude implementation of MOV* instructions | Karel Kočí | 2017-11-25 | |
| | | | | I don't like how it's implemented but I have no other idea atm. | |||
* | Implement instructions for moving from and to HI and LO registers | Karel Kočí | 2017-11-25 | |
| | ||||
* | Fix SLTU instruction | Karel Kočí | 2017-11-25 | |
| | ||||
* | Test pipelined core | Karel Kočí | 2017-11-25 | |
| | ||||
* | Implement some logical operations | Karel Kočí | 2017-11-21 | |
| | ||||
* | Implement some immediate arithmetic instructions | Karel Kočí | 2017-11-21 | |
| | ||||
* | Implement tests for few more arithmetic instructions | Karel Kočí | 2017-11-21 | |
| | ||||
* | Implement and test ADD | Karel Kočí | 2017-11-21 | |
| | ||||
* | Add possibility to compare memory and registers state | Karel Kočí | 2017-11-21 | |
| | | | | | | For core testing we want to compare whole memory and registers. Registers are pretty simple but in case of memory it is some what more complicated and required its own tests to be sure that it works. |