[Libre-soc-bugs] [Bug 806] Nest should be able to run at different clock rate than main CPU
bugzilla-daemon at libre-soc.org
bugzilla-daemon at libre-soc.org
Thu Apr 14 13:11:32 BST 2022
https://bugs.libre-soc.org/show_bug.cgi?id=806
--- Comment #16 from Luke Kenneth Casson Leighton <lkcl at lkcl.net> ---
not quite.
https://git.libre-soc.org/?p=ls2.git;a=commitdiff;h=b586cc83eea22270d7914e7315565118c18fbf75
there are 4 CLKDIVFs available and 10 ECLKSYNCBs
so it is possible to a separate spi clock domain
from dram domain from hyperram domain
Info: CLKDIVF: 1/ 4 25%
Info: ECLKSYNCB: 1/ 10 10%
i'd prefer they were done one at a time (there is a project
"minimum patch" rule), let's focus on DRAM first
dram_clk_freq. also please keep to under 80 chars
+ self.ddrphy = drs(ECP5DDRPHY(ddr_pins,
sys_clk_freq=memory_clk_freq))
ah... yyeah that's ok. for now.
yes, annoyingly:
- drambone = gramWishbone(dramcore, features={'stall'})
+ drambone = gramWishbone(dramcore)
needing two Asyncs here is annoying, removing the features={stall}
is the right thing to do, it's the WBAsyncBridge that needs it.
yes:
+ self.drambone_async_br = WBAsyncBridge(
+ master_features={'stall'})
except stall=cyc&~ack has to be applied (i took it out
of WBAsyncBridge because it will quickly become hell:
unintended consequences)
@@ -470,6 +557,7 @@ class DDR3SoC(SoC, Elaboratable):
self.memory_map = self._decoder.bus.memory_map
self.clk_freq = clk_freq
+ self.memory_clk_freq = memory_clk_freq
self.fpga = fpga
dram_clk_freq. then, later, spi_clk_freq. if they happen to be
made the same they can probably be aliased. TODO on that one.
y'know what? i'll pull this over. there's great bits of code here
in easy-to-cut-paste chunks
--
You are receiving this mail because:
You are on the CC list for the bug.
More information about the libre-soc-bugs
mailing list