Bug 528 - parameter needed in coriolis2 to specify individually which clocks are to have a Clock Tree
Summary: parameter needed in coriolis2 to specify individually which clocks are to hav...
Status: CONFIRMED
Alias: None
Product: Libre-SOC's first SoC
Classification: Unclassified
Component: Hardware Layout (show other bugs)
Version: unspecified
Hardware: PC Linux
: --- enhancement
Assignee: Jean-Paul Chaput
URL:
Depends on:
Blocks: 199 383
  Show dependency treegraph
 
Reported: 2020-11-10 13:20 GMT by Luke Kenneth Casson Leighton
Modified: 2020-11-10 16:08 GMT (History)
1 user (show)

See Also:
NLnet milestone: NLNet.2019.02.029.Coriolis2
total budget (EUR) for completion of task and all subtasks: 0
budget (EUR) for this task, excluding subtasks' budget: 0
parent task for budget allocation:
child tasks for budget allocation:
The table of payments (in EUR) for this task; TOML format:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Luke Kenneth Casson Leighton 2020-11-10 13:20:37 GMT
currently in the new niolib plugin the assumption is that all clocks which
are given separate domains (and therefore separate IO Rings) will be external
and therefore require a global H Clock Tree to cover the entire ASIC

https://bugs.libre-soc.org/show_bug.cgi?id=155#c24

if we have an external 24mhz Reference Clock, we do not actually want
this to result in a full global H-Clock Tree covering the entire ASIC.

the only place we want it to go is: to the PLL block.

however it is the *output* from the PLL that we want to have a global H-Clock
Tree.

but, given that this could run as high as 300 mhz we do *NOT* want it to
go out as an external pin.

bottom line is that we need to be able to specify, as a coriolis2/settings.py
configuration parameter, exactly which clocks are to have an H-Tree created
for them.

not for the assumption to be that if it is a clock, and it is on the IORing,
it *must* have an H-Tree.
Comment 1 Luke Kenneth Casson Leighton 2020-11-10 16:08:34 GMT
Jean-Paul:

index f1fc2f61..9877199d 100644
--- a/cumulus/src/plugins/alpha/chip/corona.py
+++ b/cumulus/src/plugins/alpha/chip/corona.py

+    @property
+    def coronaCks ( self ):
+        if self.corona.conf.useClockTree: return self.corona.conf.coronaCks
+        return []
+

this would be parameteriseable in coriolis2/settings.py (or chipConf
dictionary) to allow explicit naming of clocks that were to be an H-Tree.

in the ls180 case these would be:

* JTAG_TCK
* PLL_OUT_CLK

they would *NOT* be:

* JTAG_TCK
* PLL_OUT_CLK
* SYS_CLK

but, note: the IORing clocks would be:

* JTAG_TCK
* SYS_CLK

the H-Tree clocks *are not the same as* the IOring clocks.

this i believe is what we need.  is this easily achievable by simply
modifying corona.Side class coronaCks function to return a user-defined
list?

if so then i could try that.