!dASuPAbUCerhCYFqJs:matrix.org

nRF Rust

176 Members
Discussion medium for the nRF-rs organization on Github15 Servers

Load older messages


SenderMessageTime
2 May 2022
@dirbaio:matrix.orgdirbaioRAM is mirrored, both addrs access the RAM13:42:55
@dirbaio:matrix.orgdirbaiobut I guess since the slave is the same, both addrs trigger the errata :(13:43:07
@jamesmunns:beeper.comJames MunnsI don't need to reserve RAM regions for ICODE, right?13:58:51
@dirbaio:matrix.orgdirbaioas long as you don't use it, it's ok not to13:59:32
@jamesmunns:beeper.comJames MunnsI do13:59:38
@jamesmunns:beeper.comJames MunnsOr at least, I turn on ICACHE13:59:44
@jamesmunns:beeper.comJames Munns
        // Enable instruction caches for MAXIMUM SPEED
        device.NVMC.icachecnf.write(|w| w.cacheen().set_bit());
        cortex_m::asm::isb();
14:00:21
@dirbaio:matrix.orgdirbaiobut you don't run code out of RAM at 0x0800_0000 right?14:00:30
@jamesmunns:beeper.comJames Munnsnope14:00:36
@dirbaio:matrix.orgdirbaiothen it's OK14:00:39
@jamesmunns:beeper.comJames Munnsword14:00:42
@jamesmunns:beeper.comJames Munnswell14:00:49
@jamesmunns:beeper.comJames MunnsI might14:00:54
@jamesmunns:beeper.comJames Munnsbut that's reserved (for the application)14:01:03
@jamesmunns:beeper.comJames Munnsohhh14:01:26
@jamesmunns:beeper.comJames MunnsShould I be executing from a different location?14:01:56
@jamesmunns:beeper.comJames MunnsI load code to 0x20020000, then run from there, but I don't set the execution to 0x0802000014:02:45
@dirbaio:matrix.orgdirbaioyou can execute code from ram at 0x2000_0000 just fine. I think the only advantage for 0x0800_0000 is it's faster, because the core can do CODE and DATA fetches at the same time...?? vs if you use 0x2000_0000 for code then you force everything through DATA14:03:07
@dirbaio:matrix.orgdirbaio * you can execute code from ram at 0x2000_0000 just fine. I think the only advantage for 0x0800_0000 is it's faster, because the core can do DCODE and ICODE fetches at the same time...?? vs if you use 0x2000_0000 for code then you force everything through DCODE 14:03:50
@jamesmunns:beeper.comJames Munnshuh, interesting14:03:51
@jamesmunns:beeper.comJames MunnsI'm not doing enough "perf"y things in "userspace" yet, but I might try that later.14:04:20
@jamesmunns:beeper.comJames Munnsso icache really only affects the kernel, which is executing out of on-chip flash14:04:36
@dirbaio:matrix.orgdirbaioyep14:05:21
@jamesmunns:beeper.comJames MunnsJust a fun PSA, for all you nrf folks out there :D15:01:06
4 May 2022
@henrik_alser:matrix.orghenrik_alser changed their profile picture.18:57:58
6 May 2022
@daja_2:matrix.orgAi Maiga (dajamante) changed their profile picture.14:35:36
8 May 2022
@osvald:matrix.orgosvald joined the room.05:33:19
9 May 2022
@dkhayes117:matrix.orgdkhayes117 joined the room.16:40:19
16 May 2022
@idahotokens:matrix.org@idahotokens:matrix.org joined the room.11:25:55
@idahotokens:matrix.org@idahotokens:matrix.org left the room.16:02:47

There are no newer messages yet.


Back to Room List