Threaded Mode | Linear Mode

cyrille de brébisson · 01-29-2018, 06:58 AM

Hello,

Well, since you guys want "precise" info, I had a look at the source code.

grad are converted to deg if needed.

The calc has a flag (fix) which tells it if the result is a remarquable value (0° in all modes, n*30° or tan(45) in deg/grad).

This flag is also put on entry if 0 is detected.

The calc then does a modulo 45 or pi/2, depending, using 31 digits.
It then applies cordic, which yeild a x and y pair on the axis of the angle.
tan=y/x cos/sin are calculated from tan using 1/sqrt(1+tan()²).

Then, if the flag was set, the result is rounded.
and this is it.

Cheers,
Cyrille

rflesch · (This post was last modified: 01-29-2018 08:26 AM by rflesch.)

(01-29-2018 06:58 AM)cyrille de brébisson Wrote: Hello,

Well, since you guys want "precise" info, I had a look at the source code.

(...)

Then, if the flag was set, the result is rounded.
and this is it.

Cheers,
Cyrille

Tank you!!! From my point of view (I'm the OP) this is settled now.

emece67 · 01-29-2018, 12:31 PM

(01-26-2018 10:23 AM)Mike (Stgt) Wrote: For the inquisitive here a compare of the HP-15C firware trace from DEG 360 SIN and 720 SIN.

Code:

SUPERC - CMS FILE/LINE/WORD/BYTE/SFOR COMPARE UTILITY - V3.01 (11/22/95) 01/26/18 11.05 PAGE 1 NEW: SIN 720 C OLD: SIN 360 C LISTING OUTPUT SECTION (LINE COMPARE) ID NEW FILE LINES ID OLD FILE LINES N-LN# O-LN# ----+----1----+----2----+----3--- ----+----1----+----2----+----3--- 0648: 1EE C=C+C ALL ! 0648: 1EE C=C+C ALL 00241 00241 0649: 1EE C=C+C ALL ! 0649: 1EE C=C+C ALL 00242 00242 064A: 3CE CSL ALL ! 064A: 3CE CSL ALL 00243 00243 064B: 0EE B<>C ALL ! 064B: 0EE B<>C ALL 00244 00244 064C: 2FA ?C#0 M ! 064C: 2FA ?C#0 M 00245 00245 064D: 09F JC +13 0660 ! 064D: 09F JC +13 0660 00246 00246 064E: 18E A=A-B ALL ! 064E: 18E A=A-B ALL 00247 00247 064F: 3FB JNC -01 064E ! 064F: 3FB JNC -01 064E 00248 00248 064E: 18E A=A-B ALL ! 064E: 18E A=A-B ALL 00249 00249 064F: 3FB JNC -01 064E ! 064F: 3FB JNC -01 064E 00250 00250 I -064E: 18E A=A-B ALL ! 00251 I -064F: 3FB JNC -01 064E ! 00252 0650: 12E A=A+B ALL ! 0650: 12E A=A+B ALL 00253 00251 0651: 3EE ASL ALL ! 0651: 3EE ASL ALL 00254 00252 0652: 266 C=C-1 X ! 0652: 266 C=C-1 X 00255 00253 0653: 3DB JNC -05 064E ! 0653: 3DB JNC -05 064E 00256 00254 0654: 04E C=0 ALL ! 0654: 04E C=0 ALL 00257 00255 0655: 0EE B<>C ALL ! 0655: 0EE B<>C ALL 00258 00256 0656: 198 C=M ! 0656: 198 C=M 00259 00257 0657: 1EE C=C+C ALL ! 0657: 1EE C=C+C ALL 00260 00258 0658: 04C ?FS 4 ! 0658: 04C ?FS 4 00261 00259 0659: 01B JNC +03 065C ! 0659: 01B JNC +03 065C 00262 00260 *** CHANGE SECTION CUTOFF ****************************** SUPERC - CMS FILE/LINE/WORD/BYTE/SFOR COMPARE UTILITY - V3.01 (11/22/95) 01/26/18 11.05 PAGE 2 NEW: SIN 720 C OLD: SIN 360 C LINE COMPARE SUMMARY AND STATISTICS 1874 NUMBER OF LINE MATCHES 2 TOTAL CHANGES (PAIRED+NONPAIRED CHNG) 0 REFORMATTED LINES 0 PAIRED CHANGES (REFM+PAIRED INS/DEL) 2 NEW FILE LINE INSERTIONS 2 NON-PAIRED INSERTS 0 OLD FILE LINE DELETIONS 0 NON-PAIRED DELETES 1876 NEW FILE LINES PROCESSED 1874 OLD FILE LINES PROCESSED LISTING-TYPE = CHNG COMPARE-COLUMNS = 1:33 LONGEST-LINE = 33 PROCESS OPTIONS USED: NARROW
From pushing SIN to display of the result (to be exact: to light sleep) it takes only two more firmware instructions to compute the result for 720 instead of 360.

Ciao.....Mike

Sorry for my ignorance but, what tools have you used to generate such firmware traces?

Thanks & regards.

Massimo Gnerucci · (This post was last modified: 01-29-2018 12:53 PM by Massimo Gnerucci.)

(01-29-2018 12:31 PM)emece67 Wrote:
(01-26-2018 10:23 AM)Mike (Stgt) Wrote: For the inquisitive here a compare of the HP-15C firware trace from DEG 360 SIN and 720 SIN.

Code:

SUPERC - CMS FILE/LINE/WORD/BYTE/SFOR COMPARE UTILITY - V3.01 (11/22/95) 01/26/18 11.05 PAGE 1

Sorry for my ignorance but, what tools have you used to generate such firmware traces?

Thanks & regards.

I bet it's running on an IBM mainframe running VM/CMS or z/VM. (or an emulation on Hercules...)
Those were the days...

emece67 · 01-29-2018, 04:23 PM

(01-29-2018 02:41 PM)Mike (Stgt) Wrote: Short answer: my own tools.
Long answer: do you have time for a looong tale? (negligibly, this question is a citation from Waterhole No. 3, when Foggers asks Cole "How come..?")

...

An interesting tale, & not so looong as expected ;-)

Thanks & regards.