Re: [Chicken-hackers] [PATCH] Improve performance by skipping mutations on non-stack pointers

2014-04-06 Thread Jörg F. Wittenberger
Good catch! Though I don't understand what those counters are good for. /Jörg Am 05.04.2014 21:19, schrieb Peter Bex: Hi all, I happened to be working on some benchmarks for the Postgres egg and due to random reasons started digging into why (time) was showing mutations in the GC even if I

Re: [Chicken-hackers] [PATCH] Improve performance by skipping mutations on non-stack pointers

2014-04-06 Thread Peter Bex
On Sun, Apr 06, 2014 at 12:25:08PM +0200, Jörg F. Wittenberger wrote: Good catch! Though I don't understand what those counters are good for. Which counters? The total/tracked mutation ones? Those are useful info for when using the TIME macro to understand the performance profile of your

[Chicken-hackers] [PATCH] Improve performance by skipping mutations on non-stack pointers

2014-04-05 Thread Peter Bex
Hi all, I happened to be working on some benchmarks for the Postgres egg and due to random reasons started digging into why (time) was showing mutations in the GC even if I wasn't mutating anything. I still haven't figured that one out, but I stumbled across a nice performance boost that I