HP Forum Archive 21

[ Return to Index | Top of Index ]

WP34S: New compiler to be tested Message #1 Posted by Marcus von Cube, Germany on 31 Aug 2012, 7:37 a.m. I've compiled the WP34S binaries with a different compiler which produces more compact code. Since I'm not at home I can't test if this firmware runs at all. You'll find it on SVN in the trunk/realbuild/realbuild-cs.zip file. Is anybody willing to try this build on the hardware? It may work perfectly (with a library space one KB larger than before) or it may not work at all. Anything in between is possible, too. Thanks to Eric for pointing me to the new CodeSourcery release.

(deleted post) Message #2 Posted by deleted on 31 Aug 2012, 10:53 a.m., in response to message #1 by Marcus von Cube, Germany This Message was deleted. This empty message preserves the threading when a post with followup(s) is deleted. If all followups have been removed, the original poster may delete this post again to make this placeholder disappear.

Re: WP34S: New compiler to be tested Message #3 Posted by Marcus von Cube, Germany on 31 Aug 2012, 1:52 p.m., in response to message #2 by deleted The new build (3247) is hidden in a ZIP file as stated in my initial posting.

(deleted post) Message #4 Posted by deleted on 31 Aug 2012, 2:00 p.m., in response to message #3 by Marcus von Cube, Germany This Message was deleted. This empty message preserves the threading when a post with followup(s) is deleted. If all followups have been removed, the original poster may delete this post again to make this placeholder disappear.

Re: WP34S: New compiler to be tested Message #5 Posted by Marcus von Cube, Germany on 31 Aug 2012, 2:06 p.m., in response to message #4 by deleted It is in SVN but as long as it's not the normal build I'll put it in a different place, hence the ZIP file in the realbuild directory.

Re: WP34S: New compiler to be tested Message #6 Posted by Walter B on 31 Aug 2012, 2:47 p.m., in response to message #5 by Marcus von Cube, Germany Hmmmh, the whole SVN is for preliminary stuff, isn't it? Anyway, flashed and got 60 words less in FM plus a strange exponent in the "Free:" display of STATUS. Anything else to test?

Re: WP34S: New compiler to be tested Message #7 Posted by Marcus von Cube, Germany on 31 Aug 2012, 4:47 p.m., in response to message #6 by Walter B Last time we used the CodeSourcery compiler the solver was broken, to be precise, the x~0 test didn't return. My question is if the calculator works and if there are strange observations. Free library space should be 1 KB more than with the previous version (Yagarto compiler).

Re: WP34S: New compiler to be tested Message #8 Posted by Walter B on 31 Aug 2012, 4:59 p.m., in response to message #7 by Marcus von Cube, Germany Quote: Free library space should be 1 KB more than with the previous version (Yagarto compiler). As mentioned, this is not what I observed. I'll keep the calc.bin build 3247 on my calculator for some time and check if I'll see strange stuff - some hints where to look for such oddities, however, would be helpful.

Re: WP34S: New compiler to be tested Message #9 Posted by Don Shepherd on 31 Aug 2012, 5:20 p.m., in response to message #8 by Walter B You guys ever hear of email?

Re: WP34S: New compiler to be tested Message #10 Posted by fhub on 31 Aug 2012, 5:23 p.m., in response to message #9 by Don Shepherd Quote: You guys ever hear of email? No, they are still using carrier pigeons. ;-)

Re: WP34S: New compiler to be tested Message #11 Posted by Marcus von Cube, Germany on 1 Sept 2012, 4:43 a.m., in response to message #9 by Don Shepherd Don, I agree that we should have moved the last three or four posts to a different medium. OTH, I need people who try the new build to make sure it works (or not). The main reason to use the new compiler is to gain some more headroom for the library but not at the expense of any malfunction introduced by it. The more people try, the better I know. I can't do the tests myself right now because I'm not at home.

Re: WP34S: New compiler to be tested Message #12 Posted by Don Shepherd on 1 Sept 2012, 10:38 a.m., in response to message #11 by Marcus von Cube, Germany Thanks Marcus. I do understand the need to have people try the new compiler, but I would have thought that by now you would have developed an email list of people who are regularly using and testing the wp34s and you could send the request directly to them.

Re: WP34S: New compiler to be tested Message #13 Posted by Cristian Arezzini on 1 Sept 2012, 11:04 a.m., in response to message #12 by Don Shepherd But why? Is this still because the WP34s isn't really an official HP calculator? Still, it attracts much interest here. Or should we only post here things that will be of interest to *all* users of the forum? I don't understand it. Personally, if I see discussions with a title that doesn't interest me, I just skip them. They don't hurt me.

Re: WP34S: New compiler to be tested Message #14 Posted by Harald on 1 Sept 2012, 2:25 p.m., in response to message #12 by Don Shepherd Quote: Thanks Marcus. I do understand the need to have people try the new compiler, but I would have thought that by now you would have developed an email list of people who are regularly using and testing the wp34s and you could send the request directly to them. And I thought a forum was exactly to do things like this?!

Re: WP34S: New compiler to be tested Message #15 Posted by Reth on 2 Sept 2012, 1:50 a.m., in response to message #12 by Don Shepherd I don't see you complaining about lots of OT's posted here, neither those about calculators of other makes. WP34S is a hot issue and I appreciate the fact that the developers team keeps us posted about all of its aspects.

Re: WP34S: New compiler to be tested Message #16 Posted by Walter B on 2 Sept 2012, 2:47 a.m., in response to message #8 by Walter B The calculator needed a full RESET - there were lots of strange stuff in program memory. After resetting, the idle exponent section in STATUS still contains garbage but seems to work for new program labels set (just for the digits needed, however). And the amount of free flash memory is 5953 words with build 3247 instead of 6014 with the build before. So the advertized 1kB gain is still missing :-(

[ Return to Index | Top of Index ]

Go back to the main exhibit hall