[WP-34S] Emulator/debug - single/double copy/paste

[pascal_meheut]

(06-12-2014 07:58 AM)Martin Hepperle Wrote:  Just a small README file could save the world...

Strangely enough, it has been there for a while now and simply searching for all the files named "README something" would have helped you find the answer and navigate through "the mess".

Asking questions and for help before criticizing is also an efficient practice.

[Marcus von Cube]

(06-13-2014 10:51 PM)pito Wrote:  

(06-13-2014 08:48 PM)Marcus von Cube Wrote:  [This looks like an Eclipse artifact from an automatic scan of all C code, including the MFC based emulator code. You can safely ignore it.

Now, with makeflash.cmd (with the path set to my yagarto) I get - see flash-build.log below.

Can you post the contents of your windows32_realbuild directory? Any error messages which were only visible on the screen?

[pito]

(06-14-2014 08:29 AM)Marcus von Cube Wrote:  Can you post the contents of your windows32_realbuild directory? Any error messages which were only visible on the screen?

No visible err messages on the screen.

Code:

 D:\MyCode\Calculators\WP34s>dir
 Directory of D:\MyCode\Calculators\WP34s
06/14/2014  00:54    <DIR>          .
06/14/2014  00:54    <DIR>          ..
06/14/2014  11:02             1,436 flash-build.log
06/14/2014  00:54               807 flash-build.zip
06/14/2014  00:34               804 makeflash.cmd
06/13/2014  22:12             1,407 makelib.cmd
06/13/2014  22:12               345 makewin.cmd
06/13/2014  21:17               101 revision.txt
06/13/2014  23:32    <DIR>          trunk
               6 File(s)          4,900 bytes
               3 Dir(s)  51,382,497,280 bytes free
D:\MyCode\Calculators\WP34s>makeflash.cmd
D:\MyCode\Calculators\WP34s>cd trunk
D:\MyCode\Calculators\WP34s\trunk> cd windows32_realbuild
D:\MyCode\Calculators\WP34s\trunk\windows32_realbuild>ls
compile_consts.exe  consts  genchars7.exe  genfont.exe  obj
D:\MyCode\Calculators\WP34s\trunk\windows32_realbuild>dir
Directory of D:\MyCode\Calculators\WP34s\trunk\windows32_realbuild
06/14/2014  10:56    <DIR>          .
06/14/2014  10:56    <DIR>          ..
06/14/2014  10:56           216,237 compile_consts.exe
06/14/2014  10:56    <DIR>          consts
06/14/2014  10:56            96,857 genchars7.exe
06/14/2014  10:56           127,045 genfont.exe
06/14/2014  10:56    <DIR>          obj
               3 File(s)        440,139 bytes
               4 Dir(s)  51,383,758,848 bytes free

obj is empty, and consts is full of const_XY.c plus a Makefile (see enclosed).
When I run makeflash again, the lcdgen.exe (95k) appears too (the same err message in the log).

Make:

Code:

D:\MyCode\Calculators\WP34s>make --version
GNU Make 3.81
Copyright (C) 2006  Free Software Foundation, Inc.
This is free software; see the source for copying conditions.
There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A
PARTICULAR PURPOSE.

This program built for i686-pc-mingw32

[Marcus von Cube]

I have two remarks:

Don't use any GCC version > 4.6. We had code generation errors with GCC 4.7. So please try to get an older Yagarto build.

Your make version doesn't seem to like the -C consts option. This option makes make switch to the consts subdirectory of windows32_realbuild. Mine is a slightly newer build than yours:

Code:


C:\APPS\MinGW\bin>mingw32-make.exe --version
GNU Make 3.82
Built for i386-pc-mingw32
Copyright (C) 2010  Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.

[pito]

(06-14-2014 11:42 AM)Marcus von Cube Wrote:  I have two remarks:

Don't use any GCC version > 4.6. We had code generation errors with GCC 4.7. So please try to get an older Yagarto build.

Your make version doesn't seem to like the -C consts option.

Yea, the make was old..
I've replaced the yagarto with 4.5.2 http://sourceforge.net/projects/yagarto/.../20110318/
Now it moves forward a bit, getting errors with libraries build (probably) see below..
And the real folder:

Code:

 Directory of D:\MyCode\Calculators\WP34s\trunk\realbuild

06/14/2014  15:35    <DIR>          .
06/14/2014  15:35    <DIR>          ..
06/14/2014  15:33           663,816 calc
06/14/2014  15:33           122,624 calc.bin
06/14/2014  15:33           188,160 calc.tmp
06/13/2014  18:48           122,368 calc_full.bin
06/14/2014  15:35           687,399 calc_ir
06/14/2014  15:35           128,000 calc_ir.bin
06/14/2014  15:35           193,536 calc_ir.tmp
06/13/2014  18:48           127,488 calc_ir_full.bin
06/14/2014  15:34           672,156 calc_xtal
06/14/2014  15:34           124,416 calc_xtal.bin
06/14/2014  15:34           189,952 calc_xtal.tmp
06/13/2014  18:48           124,160 calc_xtal_full.bin
06/13/2014  18:48               151 join.cmd
06/14/2014  15:33           140,240 mapfile.txt
06/14/2014  15:35           145,224 mapfile_ir.txt
06/14/2014  15:34           141,676 mapfile_xtal.txt
06/13/2014  18:48               414 README
06/13/2014  23:10         1,111,639 realbuild.rar
06/14/2014  15:33               434 summary.txt
06/14/2014  15:35               434 summary_ir.txt
06/14/2014  15:34               434 summary_xtal.txt
06/14/2014  15:33            59,190 symbols.txt
06/14/2014  15:35            62,843 symbols_ir.txt
06/14/2014  15:34            60,704 symbols_xtal.txt
              24 File(s)      5,067,458 bytes
               2 Dir(s)  51,341,078,528 bytes free

[Marcus von Cube]

(06-14-2014 01:54 PM)pito Wrote:  I've replaced the yagarto with 4.5.2 http://sourceforge.net/projects/yagarto/.../20110318/

4.6.0 works for me and is used to create the official builds. It creates more compact code than 4.5.2.

(06-14-2014 01:54 PM)pito Wrote:  Now it moves forward a bit, getting errors with libraries build (probably) see below..

You should have checked out the complete tree. Unfortunately the library directory is in the project root, not in trunk. This is a mistake we made long ago...

[pito]

(06-14-2014 02:24 PM)Marcus von Cube Wrote:  You should have checked out the complete tree.

I did try to checkout the code above the trunk few times in past, but my svn was against. Now it checkouts the complete tree, interesting (I've upgraded my svn, maybe that has helped).

I've changed the yagarto to 460 (http://sourceforge.net/projects/yagarto/.../20110429/). Except the stuff with older make, I had to change the path to my yagarto in the makeflash.cmd, and to add the path to /tools into wp34s_asm.pl (at line 185, as recommended in a previous post). Also I work with complete tree now, so it finds the libs too..

Now I do not see any obvious error and the real build looks like (based on r3667):

Code:

 Directory of D:\MyCode\Calculators\WP34s_2\code\trunk\realbuild

06/14/2014  16:55    <DIR>          .
06/14/2014  16:55    <DIR>          ..
06/14/2014  16:54           600,607 calc
06/14/2014  16:54           118,016 calc.bin
06/14/2014  16:54           183,552 calc.tmp
06/14/2014  16:55           121,344 calc_full.bin
06/14/2014  16:54           621,205 calc_ir
06/14/2014  16:55           123,136 calc_ir.bin
06/14/2014  16:55           188,672 calc_ir.tmp
06/14/2014  16:55           126,464 calc_ir_full.bin
06/14/2014  16:54           607,781 calc_xtal
06/14/2014  16:54           119,552 calc_xtal.bin
06/14/2014  16:54           185,088 calc_xtal.tmp
06/14/2014  16:55           122,880 calc_xtal_full.bin
06/14/2014  16:37               151 join.cmd
06/14/2014  16:54           113,295 mapfile.txt
06/14/2014  16:54           115,366 mapfile_ir.txt
06/14/2014  16:54           114,473 mapfile_xtal.txt
06/14/2014  16:37               414 README
06/14/2014  16:54               434 summary.txt
06/14/2014  16:55               434 summary_ir.txt
06/14/2014  16:54               434 summary_xtal.txt
06/14/2014  16:54            62,393 symbols.txt
06/14/2014  16:55            66,061 symbols_ir.txt
06/14/2014  16:54            63,960 symbols_xtal.txt
06/14/2014  16:55             3,328 wp34s-lib.dat
              24 File(s)      3,659,040 bytes
               2 Dir(s)  50,865,963,008 bytes free

The .bin sizes are not the same as the original ones, but maybe it will work.
The only warning I've found is something like:

Code:

// WARNING: Unrecognized O/S (msys). Some features may not work as expected.

in flash-build.log..

[Marcus von Cube]

(06-14-2014 03:15 PM)pito Wrote:  The .bin sizes are not the same as the original ones, but maybe it will work.

You may have a mixture of object files from the different compilers. You can compare the map files with those on the repository. Just clean up everything and start over.

(06-14-2014 03:15 PM)pito Wrote:  The only warning I've found is something like:

Code:

// WARNING: Unrecognized O/S (msys). Some features may not work as expected.

in flash-build.log..

I can't comment on this. It is obviously related to your MinGW installation. Just flash and see what happens.

[pito]

(06-14-2014 04:32 PM)Marcus von Cube Wrote:  You may have a mixture of object files from the different compilers. You can compare the map files with those on the repository. Just clean up everything and start over.

There are 2 revs of the 460 yagarto available. That may contribute a bit of noise as well..

PS: I cleaned up the library, realbuild, windows32_realbuild and getting bins which match (the size):

Code:

 Directory of D:\MyCode\Calculators\WP34s_2\code\trunk\realbuild

06/14/2014  19:23    <DIR>          .
06/14/2014  19:23    <DIR>          ..
06/14/2014  19:21           657,035 calc
06/14/2014  19:21           119,040 calc.bin
06/14/2014  19:21           184,576 calc.tmp
06/14/2014  19:23           122,368 calc_full.bin
06/14/2014  19:22           679,997 calc_ir
06/14/2014  19:22           124,160 calc_ir.bin
06/14/2014  19:22           189,696 calc_ir.tmp
06/14/2014  19:23           127,488 calc_ir_full.bin
06/14/2014  19:22           664,905 calc_xtal
06/14/2014  19:22           120,832 calc_xtal.bin
06/14/2014  19:22           186,368 calc_xtal.tmp
06/14/2014  19:23           124,160 calc_xtal_full.bin
06/14/2014  16:37               151 join.cmd
06/14/2014  19:21           123,143 mapfile.txt
06/14/2014  19:22           126,222 mapfile_ir.txt
06/14/2014  19:22           124,279 mapfile_xtal.txt
06/14/2014  16:37               414 README
06/14/2014  19:21               434 summary.txt
06/14/2014  19:22               434 summary_ir.txt
06/14/2014  19:22               434 summary_xtal.txt
06/14/2014  19:21            57,256 symbols.txt
06/14/2014  19:22            60,785 symbols_ir.txt
06/14/2014  19:22            58,644 symbols_xtal.txt
06/14/2014  19:23             3,328 wp34s-lib.dat
              24 File(s)      3,856,149 bytes
               2 Dir(s)  50,852,651,008 bytes free

Cool!

[Paul Dale]

(06-14-2014 02:24 PM)Marcus von Cube Wrote:  This is a mistake we made long ago...

My mistake actually.


- Pauli

[pito]

(06-14-2014 10:45 PM)Paul Dale Wrote:  

(06-14-2014 02:24 PM)Marcus von Cube Wrote:  This is a mistake we made long ago...

I do not think it is a mistake, considering the library is about the "universal hprpn" math programs, which may be used for any project. That means it is easier to maintain them on a single place, as it is today. I would rename the *.wp34s to something like *.hprpn and that would justify its position outside any "trunk".