Post by thread:Making Long Calls

I realize this has come up before, and I thought I understood it - but it is not working as I thought it should. I am doing development with a 16C74A while waiting on TechTools to come out with the 16C87X family module. To make sure I understood how to make a long call (i.e. cross a 2k boundary) I wrote the following test code: device pic16c74A,xt_osc,wdt_off,protect_off org 00 goto START org 05h START bsf pclath,3 call TEST goto START org 090fh TEST retlw 0 As I understood, the call would push the entire return address to the stack, so the "retlw 0" would return to the "goto START", which it does, however, since I set the pclath bit 3 to 1 before the call, it is staying 1 on the "retlw 0" and the "goto START" is going to location 1005h rather than 0005h. If I put a "bcf pclath,3" before the "retlw 0" then the program does goto location 0005h. I thought the pclath was cleared by the retlw - is it not? Thanks, David V. Fansler Network Administrator AutoCyte, Inc. 336-222-9707 Ext. 261 spam_OUTdfanslerTakeThisOuTautocyte.com Now Showing! http://www.mindspring.com\~dfansler\ Updated September 23, 1999

At 09:53 PM 6/10/99 -0400, you wrote: It is a real pain with the 2 pages. You must set and clear the PCLATH when moving between pages. You are correct in saying that it must be cleared before doing the goto START. good luck, Peter {Quote hidden}>I realize this has come up before, and I thought I understood it - but it is >not working as I thought it should. >I am doing development with a 16C74A while waiting on TechTools to come out >with the 16C87X family module. To make sure I understood how to make a long >call (i.e. cross a 2k boundary) I wrote the following test code: >device pic16c74A,xt_osc,wdt_off,protect_off > org 00 > goto START > org 05h >START > bsf pclath,3 > call TEST > goto START > org 090fh >TEST > retlw 0 > >As I understood, the call would push the entire return address to the stack, >so the "retlw 0" would return to the "goto START", which it does, however, >since I set the pclath bit 3 to 1 before the call, it is staying 1 on the >"retlw 0" and the "goto START" is going to location 1005h rather than 0005h. >If I put a "bcf pclath,3" before the "retlw 0" then the program does goto >location 0005h. > >I thought the pclath was cleared by the retlw - is it not? >Thanks, > >David V. Fansler >Network Administrator >AutoCyte, Inc. >336-222-9707 Ext. 261 >.....dfanslerKILLspam@spam@autocyte.com >Now Showing! http://www.mindspring.com\~dfansler\ Updated September 23, 1999 >

-----Original Message----- From: Peter Grey [martechKILLspamOZEMAIL.COM.AU] Sent: Thursday, October 07, 1999 12:20 AM To: .....PICLISTKILLspam.....MITVMA.MIT.EDU Subject: Re: Making Long Calls At 09:53 PM 6/10/99 -0400, you wrote: It is a real pain with the 2 pages. You must set and clear the PCLATH when moving between pages. You are correct in saying that it must be cleared before doing the goto START. good luck, Peter Thanks Peter - you have confirmed what I was afraid of! David V. Fansler Network Administrator TriPath Imaging, Inc. 336-222-9707 Ext. 261 EraseMEdfanslerspam_OUTTakeThisOuTautocyte.com <dfanslerspam_OUTautocyte.com> Now Showing! http://www.mindspring.com/~dfansler <http://www.mindspring.com/~dfansler> Updated September 23, 1999

>since I set the pclath bit 3 to 1 before the call, it is staying 1 on the >"retlw 0" and the "goto START" is going to location 1005h rather than 0005h. That's because the RET instructions only update the program counter. They don't "fix" the bits for the next jump. Andy ================================================================== Eternity is only a heartbeat away - are you ready? Ask me how! ------------------------------------------------------------------ @spam@andyKILLspamrc-hydros.com http://www.rc-hydros.com - Race Boats KILLspamandyKILLspammontanadesign.com http://www.montanadesign.com - Electronics ==================================================================

On Wed, 6 Oct 1999 21:53:14 -0400 "Fansler, David" <RemoveMEDFANSLERTakeThisOuTAUTOCYTE.COM> writes: {Quote hidden}>I realize this has come up before, and I thought I understood it - but >it is >not working as I thought it should. >I am doing development with a 16C74A while waiting on TechTools to >come out >with the 16C87X family module. To make sure I understood how to make >a long >call (i.e. cross a 2k boundary) I wrote the following test code: >device pic16c74A,xt_osc,wdt_off,protect_off > org 00 > goto START > org 05h >START > bsf pclath,3 > call TEST > goto START > org 090fh >TEST > retlw 0 > >As I understood, the call would push the entire return address to the >stack, >so the "retlw 0" would return to the "goto START", which it does, >however, >since I set the pclath bit 3 to 1 before the call, it is staying 1 on >the >"retlw 0" and the "goto START" is going to location 1005h rather than >0005h. >If I put a "bcf pclath,3" before the "retlw 0" then the program does >goto >location 0005h. > >I thought the pclath was cleared by the retlw - is it not? >Thanks, > First, note that there is a "secret instruction" of LCALL that handles the page setting bits for you on the call. That way you don't have to keep track of them. Also, you are correct, the return or retlw instructions do not modify pclath. You can substitute an LGOTO for the goto and it should work. Harold Harold Hallikainen spamBeGoneharoldspamBeGonehallikainen.com Hallikainen & Friends, Inc. See the FCC Rules at http://hallikainen.com/FccRules and comments filed in LPFM proceeding at http://hallikainen.com/lpfm ___________________________________________________________________ Get the Internet just the way you want it. Free software, free e-mail, and free Internet access for a month! Try Juno Web: dl.http://www.juno.com/dynoget/tagj.

One other thing that the PICs (at least the 16C63/ 66) do that can bite you on long calls: The number of bits loaded from PCLATH into the PC is 2 for a GOTO or CALL, but 5 when PCL is the destination of an instruction. This silly design can get you (it did me) -- say you are executing a routine at the top of the page, then jump to a table look-up (which I usually place near the end of the page). The CALL to the table routine will work fine with no PAGESEL. But when you get there and try to do an " ADDWF PCL, F ", The PC will grab more bits from the PCLATH and jump to someplace (probably) not in the table. In other words, anyplace you use a computed GOTO, not only does the destination have to be in the same 255 address block, but PCLATH has to be set. So a PAGESEL is sometimes needed even when you are not jumping between pages. Maybe this is obvious, but I hope I can save somebody else the time of making this mistake. Also, I tried to use the LCALL and LGOTO instructions when I first started using the PICs, but they seemed to not work properly and generate more instructions than needed. Could well be operator error, but I gave them up.