Recent Posts

Topic: General / Linking AsAsm output to gcc?

As a possible solution, has anyone yet created an elf2aof application?

If not where is some good consise doc on the ELF format so I can give it a go?

Topic: Announcements / Display resolution on Pi

Using an IEEE OUI is pretty nonsensical

I agree it is a weak relationship and I hang my head with shame.

I have two Pandaboards and they have MAC addresses of 02-95-10-xx-xx-xx and 02-95-48-xx-xx-xx. Unfortunately the unique ID reported by OS_ReadSysInfo,2 bears no relationship to this (20-04-40-xx-xx-xx on the surviving machine).

Beagleboards are said to return a different unique ID each time they boot up.

My IGEPv5 returns 2E-04-5F-xx-xx-xx every time from OS_ReadSysInfo,2 but has a MAC of the form 02-98-4D-xx-xx-xx. So no joy there either.

Raspberry Pi return the MAC address via OS_ReadSysInfo,2 but with the bytes in the opposite order.

Risc PC (and other non-HAL Acorn machines) is moot as it is never recognised (by ScrHelp) by its unique ID anyway.

More thought needed…

Is there a defined relationship between unique machine ID (as returned by OS_ReadSysInfo,2) and MAC address? On some machines they are equivalent, whether low endian or high endian. On others they are not. The utility on spellings.net to read the machine’s MAC address just uses OS_ReadSysInfo,2 (and works on the Pi and Iyonix from which it needs this information).

Also what call does the internet stack use to read the MAC address (configure/network/internet/interfaces/status)? This might not help as the ‘status’ information was not available on earlier machines.

Topic: General / Linking AsAsm output to gcc?

All works now

The C code:

void PutS(char *c);

int main(void)
{
 PutS("Hello World!\n");
 return 0;
}

and the assembly:

  AREA |main|, CODE, READONLY
  EXPORT PutS

PutS
  SWI &02
  MOV R15,R14

  END

And the working compile:

*asasm t01.s -o t01.o
*gcc t01.c t01.o -o t01 -mlibscl
*t01
Hello World!
*

Original post follows
I am having trouble linking AsAsm output with GCC. With a weard error.

My test case, C source:

void PutS(char *c);

int main(void)
{
 PutS("Hello World\n");
 return 0;
}

And the assembly test case:

  AREA |main|, CODE, READONLY
  EXPORT PutS

PutS
  SWI &02
  MOV R15,R14

END

The result:

*asasm t01.s -o t01.o
SDFS::RISCOSpi.$.Projects.ctest.s.t01:9: Warning: No END found
*gcc t01.c t01.o -ot01
/SDFS::RISCOSpi.$/Apps/Development/!GCC/bin/ld: error: t01.o uses hardware FP, whereas t01 uses software FP
/SDFS::RISCOSpi.$/Apps/Development/!GCC/bin/ld: failed to merge target specific data of file t01.o
collect2: error: ld returned 1 exit status
*

I have attempted using the -fpu=SoftFPA switch with AsAsm, that did not change anything.

Topic: Announcements / Status of Zap

Making use of Procedure (space) I’ve added some patches to Zap module.

I was going to use Rick’s BreakAid – but instead used Reporter.

It looks like swi OS_HeapSort doesn’t use flags – and have given
swi OS_HeapSort32 r7 =#0 (anyone know different).

As a note -:is there a easy way of converting -
&3F,6 into &FC000000
BASIC doen’t execept the former.
ie.
BIC   R6,R6,#&3F,6; =&FC000000

Topic: General / Screen banking revisited

The big problem with screen banks on the Pi is that we don’t get very good feedback from the GPU. We don’t know how much free memory it has, and we don’t know what the size limits are for framebuffers. So to try and avoid any problems the current implementation errs on the side of caution – it will try and limit the maximum framebuffer size to 8MB, the maximum height to 4095, and the maximum number of screen banks to 7. Since you can’t tell RISC OS how many screen banks you want, the driver will allocate as many screen banks as possible while still fitting within those limits. I.e.

screen_banks = max(1, min(8MB/screen_bytes, 4095/screen_height, 7))

So for 1080p you should have three screen banks in 256 colour modes (4*1080 > 4095), two screen banks in 64K colour modes (1080p @ 16bpp is just under 4MB), but only one in 16m colour modes.

At some point I’ll have to take another look at the code and either see if there’s a better way of doing things, or start hassling the Pi firmware team about any firmware-related issues that are causing problems.

Topic: Announcements / Display resolution on Pi

the manufacturer-specific bit of the MAC address […] so far I have included:
B8-27-EB-xx-xx-xx – Raspberry Pi
2E-04-5F-xx-xx-xx – IGEPv5
00-C0-32-xx-xx-xx – Risc PC
00-09-95-xx-xx-xx – Iyonix
00-1F-7B-xx-xx-xx – i.MX6
70-B3-D5-xx-xx-xx – Titanium
02-95-xx-xx-xx-xx – Pandaboard

Using an IEEE OUI is pretty nonsensical because it’s a weak relationship to the machine. For example, 00-C0-32 is the allocation for i-Cubed because presumably you have an i-Cubed NIC? Acorn’s allocation was 00-00-A4, but then they used that for machines from A3000 era onwards (including the Risc PC).

02-95-xx-xx-xx-xx can’t be valid, because OUI’s are 24 bit, unless there are 1 billion Pandaboards out there and they genuinely bought up all 64 MA-L ranges. 2E-04-5F is unregistered.

Topic: Announcements / Display resolution on Pi

Can of worms: lid replaced.

Version 1.06 – bug squashed. Many thanks.

Topic: Announcements / Display resolution on Pi

Used Zap to look at this as it preserves line numbers just like StrongED doesn’t..

At the risk of opening up a can of worms again, if you’re editing a BASIC file you can use StrongED to chase down errors at given lines. If you set it to strip line numbers on loading and then re-save them from line 1 with an increment of 1, then any error reports using something like STR$ERL can then be tracked down using the Goto function (F5).

If you don’t fancy that, you can set the global option in StrongED not to strip the BASIC file’s own line numbers, and then manually scroll down to find it.

Topic: Community Support / Prolific PL2303 driver

I am having irql_not_less_or_equal problem in my Pc

That’s why I don’t use the Prolific adaptor on my PC.

You could try a later driver, but if your hardware isn’t genuine…

Topic: Announcements / Display resolution on Pi

Version 1.05 uploaded.

Error – Missing " at line 2430.

Used Zap to look at this as it preserves line numbers just like StrongED doesn’t..

Topic: Community Support / Prolific PL2303 driver

I am having irql_not_less_or_equal problem in my Pc.

Topic: Announcements / Display resolution on Pi

Version 1.05 uploaded.

It now recognises machines (i.e. non-Raspberry Pi or Raspberry Pi with firmware/ROM before April 2016) from their unique identifier (the manufacturer-specific bit of the MAC address) as well as from OS_Hardware. So far I have included:
B8-27-EB-xx-xx-xx – Raspberry Pi
2E-04-5F-xx-xx-xx – IGEPv5
00-C0-32-xx-xx-xx – Risc PC
00-09-95-xx-xx-xx – Iyonix
00-1F-7B-xx-xx-xx – i.MX6
70-B3-D5-xx-xx-xx – Titanium
02-95-xx-xx-xx-xx – Pandaboard
My sample size is limited so I’d like to know if there are versions of those boards which have a different manufacturer-specific MAC address than shown above.

The Beagleboard seems to return a unique ID from OS_ReadSysInfo that does not correspond to the MAC address and so that is not included. The help message over the board type explains what it found and gives the unique machine ID.

It also looks for firmware at Boot:Loader.START/ELF (for the Pi) and also at the following:
Boot:Loader.MLO (IGEPv5, Pandaboard, Beagleboard if booting from eMMC/SD card)
SDFS::4.$.!Boot.Loader.MLO (eMMc on IGEPv5)
SDFS::0.$.!Boot.Loader.MLO (SD card on IGEPv5, Pandaboard and Beagleboard)
before giving up.

Topic: Announcements / Status of Zap

Thanks Peter and Chris. I’ll have a read whilst I’m holiday this coming 7 days.

Topic: General / Screen banking revisited

I think that you have to increase the GPU memory allocation.

Try changing:
gpu_mem=64
to read:
gpu_mem=128

I hope that helps. My config text has:
gpu_mem=256
That is my setting for the RPi 2B and RPi 3B.
Though I am not completely sure that this is the fix that works or not (I have applied so many little fixes with the RPi and RISC OS since 2012, and 5 different models of RPi in the mean time).

Topic: Announcements / Status of Zap

Which is available here, if anyone fancies some light reading!

Topic: Announcements / Status of Zap

Peter: Thanks for the Phoebe info. Out of curiosity where did you garland it from?

From the “Phoebe (RISC PC II) Functional Specification” that was given to Acorn Registered Developers.

Topic: General / Screen banking revisited

Having finally upgraded my gaggle of monitors to standard 1080p widescreen models (1920 × 1080), my Basic graphics progs can no longer use screen banking for 16m or 64k colour modes. I’m stuck with 256 colours for what were previously very nice animated graphics demos, but look cr*p in 256.

Is there any way I can program the Pi to bank in 1080p 64k colours. I’ve seen some posts here and on the R-Pi forums that suggest changing settings in config/txt to assign more GPU memory, changing max pixel clock and various other settings etc. Also, has anything improved banking-memory-wise with the Pi 3, or the latest versions of RO?

Any suggestions appreciated.

Topic: Announcements / Status of Zap

GPL is one of the huge limiters that has held Open Source back.

Not really. Companies are using Open Source code since the early beginning of computing. But it was almost a pico phenomena. GPL changed that.

Now, some “companies are reluctant to use the source, and thus we miss out on the potential improvements that could have been.”

Before, most “companies use the source but NEVER publish their modifications, and thus we miss out ALL the improvements done.”

Very big difference.

Topic: Announcements / Status of Zap

If the !Zap source is to be relicensed for better distribution, may I recommend an open source license (if all previous contributors would agree), that is NOT GPL, though something that does not restrict what is linked in or how it is distrubited.

GPL is one of the huge limiters that has held Open Source back. It puts limits on how source is used that causes many companies to be reluctant to use the source, and thus we miss out on the potential improvements that could have been.

Topic: General / Faster rendering for WIMP Graphics?

@Jeffrey Lee:

Thank you. I think that setting the Window background to transparent may be the solution that I need to get this working.

Topic: Announcements / PackMan package manager 0.9.1 beta released

We’ll see if it worked tonight…

It didn’t…

Topic: General / Faster rendering for WIMP Graphics?

Which part is slow – preparing the sprite or rendering it to the screen?

If it’s the rendering to screen which is slow, it’s generally best to make sure the sprite is in the same mode as the screen. The easiest way to do this is to use OS_SpriteOp 16 to create a sprite from a portion of the screen (just a 1×1 rectangle from a corner will do) and then resize it to the size you want. That way you can rely on the OS to do the tricky bit of working out what the correct sprite mode word is.

Also, make sure you aren’t using a sprite with a mask!

If there’s no mask and the pixel formats + palettes (if applicable) match then the OS will be able to use a fast plotting routine which is basically just a memcpy form the source to destination. If there’s any pixel format conversion or palette remapping involved then it will have to go through a slower, machine-generated routine which generally only does one or two pixels at a time.

It may also help if you set the window work area background colour to “transparent” (255) to stop the Wimp from filling it. That way you can avoid any redundant overdraw when you trigger a redraw of the area containing the sprite.

Topic: Announcements / Status of Zap

“Should be applied”? Surely it is licenced in some way already,

Of course it is already licenced in some manner; however wasn’t it omitted from the SD image exactly because distribution terms and such were not evident.

and this is unlikely to be changed if you consider the number of contributors over the years.

And this is something that would need to be looked into. I just took a look at the ZapRedraw module and it contains a one line copyright message. That’s it. And?

Sure, a couple of decades ago it could just be understood that releasing the source means that you could use it – but in this day and age… :-/

I never saw any indication that Zap or parts of it are licenced under the GPL,

If one is looking to slap a label on it, then one might wonder what sort of label it would be. :-)

it is a module, and therefore it is a classic example of something to be distributed “side by side” aka “aggregated” because it is not linked to the other software.

One would say yes, but the GPL FAQ specifically forbids supplying a GPL DLL with a VB application. While the FAQ isn’t the GPL, it suggests the mindset is anti anything that isn’t GPL, so mixing could be…problematic.

Original Zap by Dom Symes was under GNU Emacs General Public Licence, which is not considered “viral”.

Can this be considered authoritative?

Topic: Announcements / Status of Zap

Does anybody have any thoughts regarding what sort of licence should be applied to Zap?

“Should be applied”? Surely it is licenced in some way already, and this is unlikely to be changed if you consider the number of contributors over the years.

I never saw any indication that Zap or parts of it are licenced under the GPL, why do you bring that up? And why would it change anything wrt ZapRedraw – it is a module, and therefore it is a classic example of something to be distributed “side by side” aka “aggregated” because it is not linked to the other software.

Original Zap by Dom Symes was under GNU Emacs General Public Licence, which is not considered “viral”.

Topic: Announcements / PackMan package manager 0.9.1 beta released

Jon,

The gccsdk/autobuilder has a program called pkgindex that builds an index from packages. It’s called from a shell script called build-index. The source is in the toolsrc subdirectory.
If you want to look at it The WebSVN url is http://www.riscos.info/websvn/listing.php?repname=gccsdk&path=%2Ftrunk%2Fautobuilder%2Ftoolsrc%2F&#a467bcd81627f39d685188f886e263986.

PackMan doesn’t mind about the file name on the package url, so you can use what you like.
The convention for the package zip files in <PackageName>_<version>-<package_version> e.g. PackMan_0.90-1, but this isn’t critical.

I’ve been wishing for packages of your software for quite a while, but haven’t had time to talk to you about it. If you need any help, please let me know – my email is in the PackMan !Help file.