Hi, Thanks for the responses.
> TRAPTRACE allocates some physically contiguous memory for each cpu, > so that we can write out records using physical memory accesses > (for obvious reasons during TLB miss traps). It also sets > up a virtual mappings of the traptrace buffer, and it is > through those that we can retrieve records so as to use > them in ::trace etc. Currently we're trying to read the traptrace buffer in clock(), Our hope is that if we remove all the TRAPTRACE logging other than the TLB misses, we can catch enough of the misses just by periodically dumping the entire buffer to a file. The problem we're running into is that the virtual address seems uninitialized. We are able to access the trap_trace_ctl headers in clock(), but the virtual address pointer is set to NULL. We're unsure if the virtual address are only initialized in the kernel module debugger (or elsewhere), or if we're trying to access the virtual address of the buffer incorrectly. Alternatively, is there some way to write a daemon that periodically loads the kernel module debugger and access the ::ttrace information there? Thanks! TJ -- This message posted from opensolaris.org _______________________________________________ opensolaris-code mailing list opensolaris-code@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/opensolaris-code
