Just tried "--max-allocs-per-node 10000", compiling time increase to about 8 minutes from 2 minutes. Code size results:
code0 data code1 code2 code3 code4 code5 code6 code7 code 1C8D 1A90 4661 525B 47A7 50A5 55D0 4354 5625 1BB0 // ar168g sip 0.55.006 sdcc 3.0.1 1BFC 1A90 476C 5042 489D 5213 5612 452D 57DC 1B7E // ar168g sip 0.55.008 sdcc 3.0.6 #6969 1BB4 1A90 45EE 4ECA 4630 51D8 52A4 4A17 5310 1B65 // ar168g sip 0.55.009 sdcc 3.0.6 #6969 --max-allocs-per-node 10000 Compared with default --max-allocs-per-node, all code except code6 decreased. "code6" is most about md5 calculation. Woody http://palmmicro.com/woody/ ----- Original Message ----- From: "Lin Rongrong" <wo...@palmmicro.com> To: <sdcc-user@lists.sourceforge.net> Sent: Sunday, October 16, 2011 12:29 PM Subject: Z80 code size and compile time > Philipp, > > I tested current build #6969, all fine so far. And I noticed that the > compile time was almost the same as 3.0.1 now, and the code size vary on > different code: > > code0 data code1 code2 code3 code4 code5 code6 code7 code > 1C8D 1A90 4661 525B 47A7 50A5 55D0 4354 5625 1BB0 // ar168g sip > 0.55.006 sdcc 3.0.1 > 1BFC 1A90 476C 5042 489D 5213 5612 452D 57DC 1B7E // ar168g sip > 0.55.008 sdcc 3.0.6 #6969 > > I am using command line like 'C:\SDCC\BIN\sdcc > sipr.c -mz80 -c --std-c99 --codeseg CODE7", any suggestion for the > options of the new register allocator to reduce code size? > > Woody > > http://palmmicro.com/woody/ > ------------------------------------------------------------------------------ All the data continuously generated in your IT infrastructure contains a definitive record of customers, application performance, security threats, fraudulent activity and more. Splunk takes this data and makes sense of it. Business sense. IT sense. Common sense. http://p.sf.net/sfu/splunk-d2d-oct _______________________________________________ Sdcc-user mailing list Sdcc-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/sdcc-user
