Quick Jump Daily Digest

**_dysfunctional** · 10-22-2008, 07:03 PM

You aren't linking the source file? (Adding "objlist.o" to your OBJS line, if you are using a PSP makefile.)

s/objlist.o/what_ever_source_file_is_ including_the_blitimage_f ile.o

**Insert_Witty_Name** · 10-23-2008, 01:00 AM

I see two prototypes for blitImage() but only one actual function.

Where is the function that matches the prototype?

Code:

void blitImage(int x, int y, SDL_Surface* source, SDL_Surface* destination);

**yaustar** · 10-23-2008, 01:17 AM

You are using default parameters incorrectly:

header

Code:

void blitImage(int x, int y, SDL_Surface* source, SDL_Surface* destination SDL_Rect* sprite = NULL);

Source

Code:

void blitImage(int x, int y, SDL_Surface* source, SDL_Surface* destination, SDL_Rect* sprite)
{
	SDL_Rect coord;
	
	coord.x = x;
	coord.y = y;
	
	SDL_BlitSurface(source, sprite, destination, &coord);
}

**Auraomega** · 10-23-2008, 02:59 AM

Zitat von _dysfunctional

You aren't linking the source file? (Adding "objlist.o" to your OBJS line, if you are using a PSP makefile.)

s/objlist.o/what_ever_source_file_is_ including_the_blitimage_f ile.o

Nope, its added.

Zitat von yaustar

You are using default parameters incorrectly:

No luck there either, I'm getting an ambiguity issue:

source/objlist.cpp:23: note: candidates are: void blitImage(int, int, SDL_Surface*, SDL_Surface*)
note: void blitImage(int, int, SDL_Surface*, SDL_Surface*, SDL_Rect*)

I'm going to Google default parameters and hope something comes up (I'll edit the post if it does).

-Aura

Durrrr, brains not functioning. I had void blitImage(int x, int y, SDL_Surface* source, SDL_Surface* destination); at the top of objlist from last night when I was trying to solve the problem, this was causing the ambiguity because I forgot to remove it. All fixed and working now.

**Mr305** · 10-24-2008, 02:37 PM

What are the analog offsets ? LX, LY: 0 to 255. Right? 115 - 128 Center? :)
[I wish not to code and test that out

]

Dont you all like using webbrowser with your right thumb? ;)

**Insert_Witty_Name** · 10-24-2008, 03:27 PM

It's an unsigned char type, so 0 - 255 yes.

**homemister91** · 10-25-2008, 10:14 PM

Hey all,
would some one be able to give an example of the sceIoReadAsync() function from the sdk. I have been looking all over the forums and even in PPA source code. But i can't seem to get it to work.

Regards
Homemister

**chrisp6825** · 11-09-2008, 05:45 PM

Hi, this is still the latest tool chain release, right?
http://ps2dev.org/psp/Tools/Toolchain - 06/26/2007

**NoEffex** · 11-09-2008, 06:37 PM

Zitat von chrisp6825

Hi, this is still the latest tool chain release, right?
http://ps2dev.org/psp/Tools/Toolchain - 06/26/2007

I'm pretty sure it's the latest toolchain, but not latest sdk from the svn.

svn co svn://svn.ps2dev.org/psp/trunk/pspsdk

**Insert_Witty_Name** · 11-10-2008, 01:52 AM

It's not the latest toolchain.

Check the latest version out from the SVN.

Code:

svn co svn://svn.ps2dev.org/psp/trunk/psptoolchain

**chrisp6825** · 11-10-2008, 07:31 PM

Well.. I updated my sdk and when I was installing it, I got this..

Spoiler for Error:

And I tried searching around about it and found a post from another site where a guy had the same problem. It was pretty recent.

Anyone have any ideas?

**Art** · 11-11-2008, 06:28 PM

Hi Guys,
I have two questions,
What do the last two values represent in this:

Code:

PSP_MODULE_INFO("Program_Title", 0, 1, 0);

and more importantly,
What is the fastest way to split a short into two bytes (u8 variables).
Cheers, Art.

**pspersanonymous** · 11-11-2008, 07:44 PM

Zitat von Art

Hi Guys,
I have two questions,
What do the last two values represent in this:

Code:

PSP_MODULE_INFO("Program_Title", 0, 1, 0);

and more importantly,
What is the fastest way to split a short into two bytes (u8 variables).
Cheers, Art.

IIRC they are the version number (major and minor version). Though they can be arbitrarily set (I think they are 'char's so upto 0-255 I guess) but better check the pspsdk source for the actual macro definition.

Fastest way to split? Well the most direct way is:

Code:

u16 x = 0xABCD;
u8 a = x & 0xFF; // 0xCD
u8 b = x >> 8; // 0xAB
u16 y = (b<<8) | a; // to get back original 0xABCD

Alternatively you could do:

Code:

u16 x = 0xABCD;
u8 a = *((u8*)&x+0); // 0xCD
u8 b = *((u8*)&x+1); // 0xAB

// or written another way

a = ((u8*)&x)[0]; // 0xCD
b = ((u8*)&x)[1]; // 0xAB

I think this would be quicker since it would just be 2 load instructions, no other arithmetic manipulations.

**Art** · 11-11-2008, 10:39 PM

Thankyou :)
That was fast.

**Insert_Witty_Name** · 11-12-2008, 02:54 AM

chrisp6825: I commited a change to the SDK last night to fix the issue you mention.

Try re-installing and it should work now.

**Mr305** · 11-12-2008, 05:29 AM

Wouldn't this work too?

(0xABCD & 0x00ff) -> 0xCD; (or may be 0x00CD)
(0xABCD & 0xff00) -> 0xAB; (or may be 0xAB00)

Zitat von pspersanonymous

IIRC they are the version number (major and minor version). Though they can be arbitrarily set (I think they are 'char's so upto 0-255 I guess) but better check the pspsdk source for the actual macro definition.

Fastest way to split? Well the most direct way is:

Code:

u16 x = 0xABCD;
u8 a = x & 0xFF; // 0xCD
u8 b = x >> 8; // 0xAB
u16 y = (b<<8) | a; // to get back original 0xABCD

Alternatively you could do:

Code:

u16 x = 0xABCD;
u8 a = *((u8*)&x+0); // 0xCD
u8 b = *((u8*)&x+1); // 0xAB

// or written another way

a = ((u8*)&x)[0]; // 0xCD
b = ((u8*)&x)[1]; // 0xAB

I think this would be quicker since it would just be 2 load instructions, no other arithmetic manipulations.

**_df_** · 11-12-2008, 06:09 AM

Zitat von pspersanonymous

Code:

u16 x = 0xABCD;
u8 a = *((u8*)&x+0); // 0xCD
u8 b = *((u8*)&x+1); // 0xAB

// or written another way

a = ((u8*)&x)[0]; // 0xCD
b = ((u8*)&x)[1]; // 0xAB

Dont do this, its not portable. Byte ordering changes between platforms.

juse use the

Code:

lo = x & 0xFF;
hi = (x>>8) & 0xFF;

method instead.

**Auraomega** · 11-12-2008, 11:09 AM

Is there a data type to hold 12 digit numbers? Best thing I've come up with so far is:

Code:

unsigned long int part1 = 0;
unsigned long int part2 = 0;

...

if(part1 == ulMax)
{
     part2++;
     part1 = 0;
}
else
{
     part1++;
}

Its untidy but its only being used for a one shot application, just looking for the quickest method as its going to take a few hours to run as it is.

-Aura

EDIT:

Nevermind, I found how to use the u64 data type, for anyone wondering:
#define u64 unsigned long long int.

EDIT 2:

Question is still open, u64 won't allow me to place the value in that I require.

**pspersanonymous** · 11-12-2008, 02:02 PM

Zitat von Mr305

Wouldn't this work too?

(0xABCD & 0x00ff) -> 0xCD; (or may be 0x00CD)
(0xABCD & 0xff00) -> 0xAB; (or may be 0xAB00)

No (0xABCD & 0xff00) gives 0xAB00, if you assigned that to a char variable it would truncate to 0x00. In fact you don't really need the & 0xFF part:

Code:

u8 a = 0xABCD; // would just give 0xCD, but compiler probably will warn

// so while this works perfectly fine:
u16 x = 0xABCD;
u8 a = x; // will truncate to 0xCD
u8 b = x>>8;

// it's probably better practice to do:
u16 x = 0xABCD;
u8 a = x & 0xFF;
u8 b = (x>>8) & 0xFF;

Zitat von Auraomega

Is there a data type to hold 12 digit numbers? Best thing I've come up with so far is:

Code:

unsigned long int part1 = 0;
unsigned long int part2 = 0;

...

if(part1 == ulMax)
{
     part2++;
     part1 = 0;
}
else
{
     part1++;
}

Its untidy but its only being used for a one shot application, just looking for the quickest method as its going to take a few hours to run as it is.

-Aura

EDIT:

Nevermind, I found how to use the u64 data type, for anyone wondering:
#define u64 unsigned long long int.

EDIT 2:

Question is still open, u64 won't allow me to place the value in that I require.

What do you mean by "12 digits" ?? 12 digits in hex or in decimal ??

The u64 datatype is 64bits long (ie. 8 bytes long) so it stores 8 'char' values eg. 0x1122334455667788. To use 'long long' constant values you need to append 'll' to the end of the value:

Code:

u64 var = 0x1122334455667788ll;
printf("%llx", var);

**Davee** · 11-12-2008, 03:08 PM

You can also do,

char p = *(u8 *)&x;
char q = *(u8 *)(&x + 1);

or, you can do,

char p = _lb(&x);
char q = _lb(&x + 1);

**pspersanonymous** · 11-12-2008, 07:33 PM

Zitat von Davee

You can also do,

char p = *(u8 *)&x;
char q = *(u8 *)(&x + 1);

or, you can do,

char p = _lb(&x);
char q = _lb(&x + 1);

No no, that will give you invalid results. Note that's not the same thing as what I wrote (there's a slight but extremely important difference). Here's what I wrote originally in the above post:

Code:

u16 x = 0xABCD;
u8 a = *((u8*)&x+0); // 0xCD
u8 b = *((u8*)&x+1); // 0xAB

Notice the brackets, what you wrote:

Code:

char q = *(u8*)(&x+1); // dangerous, this will read out of bounds

here, x is a u16 datatype so the +1 doesnt add 1 to the address it actually adds 2, if x was an int, +1 would add 4 (or however long the datatype is on your platform).

So if x was stored at address 0x1000, your example would read a char from 0x1002, instead of reading from 0x1001 (ignore the real-world for a moment where stack values grow down from high-to-low addresses).

You need to watch where you put your brackets, it can make a huge difference (a matter of running correctly or crashing !!). You can try your example out, I don't think it will crash the app, but it definitely won't give you the correct result.

The second example, well, that is dependant on, firstly, running on a MIPS platform, and secondly, having those macro's defined in your setup. In that case, better off writing in pure asm (or inline asm).

**chrisp6825** · 11-12-2008, 08:15 PM

Zitat von Insert_Witty_Name

chrisp6825: I commited a change to the SDK last night to fix the issue you mention.

Try re-installing and it should work now.

Sweet thanks. Ill try it out tonight.

-=Double Post Merge =-

Hmm, I got the error again..

Spoiler for error:

If you want to see my time.h file that it's saying has the previous decleration then here it is.. (this is the one at "usr\local\pspdev\psp\inc lude\sys\"

Spoiler for time.h:

Code:

/* time.h -- An implementation of the standard Unix <sys/time.h> file.
   Written by Geoffrey Noer <[email protected]>
   Public domain; no rights reserved. */

#ifndef _SYS_TIME_H_
#define _SYS_TIME_H_

#include <_ansi.h>
#include <sys/types.h>

#ifdef __cplusplus
extern "C" {
#endif

#ifndef _WINSOCK_H
struct timeval {
  time_t      tv_sec;
  suseconds_t tv_usec;
};

struct timezone {
  int tz_minuteswest;
  int tz_dsttime;
};

#ifdef __CYGWIN__
#include <cygwin/sys_time.h>
#endif /* __CYGWIN__ */

#endif /* _WINSOCK_H */

#define ITIMER_REAL     0
#define ITIMER_VIRTUAL  1
#define ITIMER_PROF     2

struct  itimerval {
  struct  timeval it_interval;
  struct  timeval it_value;
};

/* BSD time macros used by RTEMS code */
#if defined (__rtems__) || defined (__CYGWIN__)

/* Convenience macros for operations on timevals.
   NOTE: `timercmp' does not work for >= or <=.  */
#define	timerisset(tvp)		((tvp)->tv_sec || (tvp)->tv_usec)
#define	timerclear(tvp)		((tvp)->tv_sec = (tvp)->tv_usec = 0)
#define	timercmp(a, b, CMP) 						      \
  (((a)->tv_sec == (b)->tv_sec) ? 					      \
   ((a)->tv_usec CMP (b)->tv_usec) : 					      \
   ((a)->tv_sec CMP (b)->tv_sec))
#define	timeradd(a, b, result)						      \
  do {									      \
    (result)->tv_sec = (a)->tv_sec + (b)->tv_sec;			      \
    (result)->tv_usec = (a)->tv_usec + (b)->tv_usec;			      \
    if ((result)->tv_usec >= 1000000)					      \
      {									      \
	++(result)->tv_sec;						      \
	(result)->tv_usec -= 1000000;					      \
      }									      \
  } while (0)
#define	timersub(a, b, result)						      \
  do {									      \
    (result)->tv_sec = (a)->tv_sec - (b)->tv_sec;			      \
    (result)->tv_usec = (a)->tv_usec - (b)->tv_usec;			      \
    if ((result)->tv_usec < 0) {					      \
      --(result)->tv_sec;						      \
      (result)->tv_usec += 1000000;					      \
    }									      \
  } while (0)
#endif /* defined (__rtems__) || defined (__CYGWIN__) */

int _EXFUN(gettimeofday, (struct timeval *__p, struct timezone *__z));
int _EXFUN(settimeofday, (const struct timeval *, const struct timezone *));
int _EXFUN(utimes, (const char *__path, const struct timeval *__tvp));
int _EXFUN(getitimer, (int __which, struct itimerval *__value));
int _EXFUN(setitimer, (int __which, const struct itimerval *__value,
					struct itimerval *__ovalue));

#ifdef __cplusplus
}
#endif
#endif /* _SYS_TIME_H_ */

Thanks for the help.

**_df_** · 11-13-2008, 04:57 AM

Zitat von Auraomega

Is there a data type to hold 12 digit numbers? Best thing I've come up with so far is:
Nevermind, I found how to use the u64 data type, for anyone wondering:
#define u64 unsigned long long int.

aurgh. dont do this. this is what

Code:

#include <stdint.h>

is for.

just use uint64_t

**Insert_Witty_Name** · 11-13-2008, 07:23 AM

chrisp6825: Try again from scratch, the full toolchain. I made another commit after that one.

**Hideous** · 11-13-2008, 11:29 AM

Not really C++-related, but still:

Both when using OSlib and SDL, I get massive ghosting when moving images. It's extreme, really. I'm drawing a dark green image, and moving it, and it has a black trail after it. Anyone know what's up?

**slicer4ever** · 11-13-2008, 01:44 PM

your not clearing the screen it sounds like, i'd suggest doing:

Code:

...
oslStartDrawing();
oslDrawFillRect(0,0,480,272, RGBA(0,0,0,255));
...

that should clear up your problem

**Hideous** · 11-13-2008, 01:50 PM

Nope, I am clearing it. Turns out it's the crappy screen of the PSP.

**Davee** · 11-13-2008, 02:01 PM

Zitat von pspersanonymous

No no, that will give you invalid results. Note that's not the same thing as what I wrote (there's a slight but extremely important difference). Here's what I wrote originally in the above post:

Code:

u16 x = 0xABCD;
u8 a = *((u8*)&x+0); // 0xCD
u8 b = *((u8*)&x+1); // 0xAB

Notice the brackets, what you wrote:

Code:

char q = *(u8*)(&x+1); // dangerous, this will read out of bounds

here, x is a u16 datatype so the +1 doesnt add 1 to the address it actually adds 2, if x was an int, +1 would add 4 (or however long the datatype is on your platform).

So if x was stored at address 0x1000, your example would read a char from 0x1002, instead of reading from 0x1001 (ignore the real-world for a moment where stack values grow down from high-to-low addresses).

You need to watch where you put your brackets, it can make a huge difference (a matter of running correctly or crashing !!). You can try your example out, I don't think it will crash the app, but it definitely won't give you the correct result.

The second example, well, that is dependant on, firstly, running on a MIPS platform, and secondly, having those macro's defined in your setup. In that case, better off writing in pure asm (or inline asm).

Damn, what the hell was I thinking? xD

Thanks for that.

**slicer4ever** · 11-13-2008, 02:17 PM

Zitat von Hideous

Nope, I am clearing it. Turns out it's the crappy screen of the PSP.

hmm, than i'm not sure what would cause such a huge ghosting

**yaustar** · 11-14-2008, 02:38 AM

Zitat von slicer4ever

hmm, than i'm not sure what would cause such a huge ghosting

High contrast images moving across the screen. It is a hardware issue with the older PSPs.

Quick Jump Daily Digest

C/C++ Programming Help Thread

LinkBack

Themen-Optionen

Tags for this Thread

Forumregeln

New to forum

Pubg

CPU Fan

Gaming Laptop vs. Rigs

How to Transfer Photos from...

Super Mario Fans Out There!

the absolutely best antivirus?

Hi Everyone