*** openpowerbot <openpowerbot!~openpower@94-226-188-34.access.telenet.be> has quit IRC | 15:04 | |
*** openpowerbot <openpowerbot!~openpower@94-226-188-34.access.telenet.be> has joined #microwatt | 15:11 | |
* openpowerbot [mattermost] <lkcl> just added mis-aligned LD to libresoc, and for lookups that go via the MMU i encountered something anomalous / strange in microwatt when SECOND_REQ is activated | 19:03 | |
openpowerbot | [mattermost] <lkcl> mmu.bin test 5 is the LD-test that deliberately crosses a page-boundary, from 0x39fffd to 0x3a0000 | 19:04 |
---|---|---|
openpowerbot | [mattermost] <lkcl> howeverrrr... the page-fault in DAR is returned as 0x3a0000 whereas in libre-soc it's 0xa30005 | 19:05 |
openpowerbot | [mattermost] <lkcl> and the reason is because the address given to SECOND_REQ in loadstore1.vhdl is aligned to the nearest dword, whereas in libre-soc's loadstore1.py i simply add 8 to the address (0x39fffd) | 19:06 |
openpowerbot | [mattermost] <lkcl> this has some... interesting consequences in dcache | 19:06 |
openpowerbot | [mattermost] <lkcl> (anywhere the 3 LSBs are used in hash lookups etc.) | 19:07 |
openpowerbot | [mattermost] <lkcl> i'm not sure what the best thing to do is, here | 19:08 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!