Hi Conor, > -----Original Message----- > From: Conor Dooley <conor@xxxxxxxxxx> > Sent: Saturday, October 14, 2023 6:56 PM > To: Linu Cherian <lcherian@xxxxxxxxxxx> > Cc: Mike Leach <mike.leach@xxxxxxxxxx>; suzuki.poulose@xxxxxxx; > james.clark@xxxxxxx; leo.yan@xxxxxxxxxx; linux-arm- > kernel@xxxxxxxxxxxxxxxxxxx; coresight@xxxxxxxxxxxxxxxx; linux- > kernel@xxxxxxxxxxxxxxx; robh+dt@xxxxxxxxxx; > krzysztof.kozlowski+dt@xxxxxxxxxx; conor+dt@xxxxxxxxxx; > devicetree@xxxxxxxxxxxxxxx; Sunil Kovvuri Goutham > <sgoutham@xxxxxxxxxxx>; George Cherian <gcherian@xxxxxxxxxxx> > Subject: Re: [EXT] Re: [PATCH 1/7] dt-bindings: arm: coresight-tmc: Add > "memory-region" property > > On Sat, Oct 14, 2023 at 11:36:37AM +0000, Linu Cherian wrote: > > > > + - description: Reserved meta data memory. Used for ETR and ETF > sinks. > > > > + > > > > + memory-region-names: > > > > + items: > > > > + - const: trace-mem > > > > + - const: metadata-mem > > > > + > > > > > > Is there a constraint required here? If we are using the memory area > > > for trace in a panic situation, then we must have the meta data > > > memory area defined? > > > Perhaps a set of names such as "etr-trace-mem", "panic-trace-mem" , > > > "panic-metadata-mem", were the first is for general ETR trace in > > > non-panic situation and then constrain the "panic-" areas to appear > together. > > > The "etr-trace-mem", "panic-trace-mem" could easily point to the > > > same area. > > > > > As noted above, we do not have other generic use case for these reserved > regions now. > > Hence two regions/names, panic-trace-mem and panic-metadata-mem > with > > constraints kept as > > minItems: 2 and maxItems: 2 would suffice ? > > Whatever you do, please delete the -mem suffix. Ack.
