HOME button doesn't work on custom firmware... Why ?

pegasus2000 · Post by **pegasus2000** » Sat Dec 29, 2007 3:09 am

When you use sceCtrlPeekBufferPositive on custom firmware, the
HOME button isn't revealed: the routine returns 0.

Do you know what is the origin of the trouble ?

sakya · Post by **sakya** » Sat Dec 29, 2007 3:19 am

Hi! :)

pegasus2000 wrote:When you use sceCtrlPeekBufferPositive on custom firmware, the
HOME button isn't revealed: the routine returns 0.

Do you know what is the origin of the trouble ?

To read buttons such as HOME, VOL_UP, NOTE, ecc.. you need to call sceCtrlPeekBufferPositive in kernel mode (in a kernel mode prx). ;)

Ciaooo
Sakya

pegasus2000 · Post by **pegasus2000** » Sat Dec 29, 2007 3:36 am

sakya wrote:Hi! :)
pegasus2000 wrote:When you use sceCtrlPeekBufferPositive on custom firmware, the
HOME button isn't revealed: the routine returns 0.

Do you know what is the origin of the trouble ?
To read buttons such as HOME, VOL_UP, NOTE, ecc.. you need to call sceCtrlPeekBufferPositive in kernel mode (in a kernel mode prx). ;)

Ciaooo
Sakya

The trouble isn't here. I'm already in a kernel mode thread (not
launched by PRX, but from main code), and the other special
keys (VOL_UP, NOTE etc) works. Only HOME button is not
detected.....

sakya · Post by **sakya** » Sat Dec 29, 2007 4:01 am

Ciao! :)

pegasus2000 wrote:The trouble isn't here. I'm already in a kernel mode thread (not
launched by PRX, but from main code), and the other special
keys (VOL_UP, NOTE etc) works. Only HOME button is not
detected.....

Strange...I use sceCtrlReadBufferPositive in a kernel mode prx and have no problem in reading PSP_CTRL_HOME.
Is it possible that sceCtrlPeekBufferPositive for some reason doesen't read the HOME button?

Code: Select all

void readButtons&#40;SceCtrlData *pad_data, int count&#41;&#123;
    k1 = pspSdkSetK1&#40;0&#41;;
    if &#40;sceKernelDevkitVersion&#40;&#41; < 0x03070110&#41;
       sceCtrlReadBufferPositive&#40;pad_data, count&#41;;
    else
       sceCtrlReadBufferPositive371&#40;pad_data, count&#41;;
    pspSdkSetK1&#40;k1&#41;;
&#125;

using this nid:

Code: Select all

	.set noreorder

#include "pspimport.s"

	IMPORT_START "sceCtrl_driver",0x00010000
	IMPORT_FUNC	"sceCtrl_driver",0x6D74BF08,sceCtrlSetSamplingCycle371 
	IMPORT_FUNC	"sceCtrl_driver",0x6A2774F3,sceCtrlSetSamplingCycle
	
	IMPORT_FUNC	"sceCtrl_driver",0x28E71A16,sceCtrlSetSamplingMode371
	IMPORT_FUNC	"sceCtrl_driver",0x1F4011E6,sceCtrlSetSamplingMode

	IMPORT_FUNC	"sceCtrl_driver",0x454455AC,sceCtrlReadBufferPositive371
	IMPORT_FUNC	"sceCtrl_driver",0x1F803938,sceCtrlReadBufferPositive

Ciaooo
Sakya

SamuraiX · Post by **SamuraiX** » Sat Dec 29, 2007 8:13 am

Have a look at my ctrlhook.prx source and you will see how its done. It was based off homehook.prx but I went to the extreme.....lol

http://openbor.svn.sourceforge.net/view ... rlhookprx/

Have fun!

pegasus2000 · Post by **pegasus2000** » Sat Dec 29, 2007 9:37 am

sakya wrote:Ciao! :)
pegasus2000 wrote:The trouble isn't here. I'm already in a kernel mode thread (not
launched by PRX, but from main code), and the other special
keys (VOL_UP, NOTE etc) works. Only HOME button is not
detected.....
Strange...I use sceCtrlReadBufferPositive in a kernel mode prx and have no problem in reading PSP_CTRL_HOME.
Is it possible that sceCtrlPeekBufferPositive for some reason doesen't read the HOME button?
Code: Select all
void readButtons&#40;SceCtrlData *pad_data, int count&#41;&#123;
    k1 = pspSdkSetK1&#40;0&#41;;
    if &#40;sceKernelDevkitVersion&#40;&#41; < 0x03070110&#41;
       sceCtrlReadBufferPositive&#40;pad_data, count&#41;;
    else
       sceCtrlReadBufferPositive371&#40;pad_data, count&#41;;
    pspSdkSetK1&#40;k1&#41;;
&#125;
using this nid:
Code: Select all
	.set noreorder

#include "pspimport.s"

	IMPORT_START "sceCtrl_driver",0x00010000
	IMPORT_FUNC	"sceCtrl_driver",0x6D74BF08,sceCtrlSetSamplingCycle371 
	IMPORT_FUNC	"sceCtrl_driver",0x6A2774F3,sceCtrlSetSamplingCycle
	
	IMPORT_FUNC	"sceCtrl_driver",0x28E71A16,sceCtrlSetSamplingMode371
	IMPORT_FUNC	"sceCtrl_driver",0x1F4011E6,sceCtrlSetSamplingMode

	IMPORT_FUNC	"sceCtrl_driver",0x454455AC,sceCtrlReadBufferPositive371
	IMPORT_FUNC	"sceCtrl_driver",0x1F803938,sceCtrlReadBufferPositive
Ciaooo
Sakya

SOLVED. I have created a dedicated function in kernel extender...

Grazie Paolo....

J.F. · Post by **J.F.** » Sat Dec 29, 2007 5:12 pm

SamuraiX wrote:Have a look at my ctrlhook.prx source and you will see how its done. It was based off homehook.prx but I went to the extreme.....lol

http://openbor.svn.sourceforge.net/view ... rlhookprx/

Have fun!

Just curious... why do you do every single button separately instead of all at once? You have more than a dozen operations when only one is needed.

adrahil · Post by **adrahil** » Sat Dec 29, 2007 10:28 pm

I don't think that "a=b & (x | y | z)" is much better than "a= b&x; a|=b&y; a|=b&z;"... I think it's even the opposite... Especially since a and b are assigned registers ;)

Mihawk · Post by **Mihawk** » Sat Dec 29, 2007 11:06 pm

adrahil wrote:I don't think that "a=b & (x | y | z)" is much better than "a= b&x; a|=b&y; a|=b&z;"... I think it's even the opposite... Especially since a and b are assigned registers ;)

If you have three constants or'ed together it's combined to just one constant at compile time. So why would it be better to do a "and plus or" operation three times instead of just one "and" operation? And what difference would it make if you're doing it with assigned registers?

Ok, if it was sarcasm, then you need to put more ";)" there ;)

hlide · Post by **hlide** » Sun Dec 30, 2007 4:02 am

ok, let's be sarcastic !

is

Code: Select all

ctrl_button |= paddata.Buttons & PSP_CTRL_SELECT;
ctrl_button |= paddata.Buttons & PSP_CTRL_START;
ctrl_button |= paddata.Buttons & PSP_CTRL_UP;
ctrl_button |= paddata.Buttons & PSP_CTRL_RIGHT;
ctrl_button |= paddata.Buttons & PSP_CTRL_DOWN;
ctrl_button |= paddata.Buttons & PSP_CTRL_LEFT;
ctrl_button |= paddata.Buttons & PSP_CTRL_LTRIGGER;
ctrl_button |= paddata.Buttons & PSP_CTRL_RTRIGGER;
ctrl_button |= paddata.Buttons & PSP_CTRL_TRIANGLE;
ctrl_button |= paddata.Buttons & PSP_CTRL_CIRCLE;
ctrl_button |= paddata.Buttons & PSP_CTRL_CROSS;
ctrl_button |= paddata.Buttons & PSP_CTRL_SQUARE;
ctrl_button |= paddata.Buttons & PSP_CTRL_HOME;
ctrl_button |= paddata.Buttons & PSP_CTRL_HOLD;
ctrl_button |= paddata.Buttons & PSP_CTRL_NOTE;
ctrl_button |= paddata.Buttons & PSP_CTRL_SCREEN;
ctrl_button |= paddata.Buttons & PSP_CTRL_VOLUP;
ctrl_button |= paddata.Buttons & PSP_CTRL_VOLDOWN;

not similar to

Code: Select all

ctrl_button = paddata.Buttons

?

it seems everything is here.

SamuraiX · Post by **SamuraiX** » Sun Dec 30, 2007 7:16 am

I know, I know.... shame on me!!!!! Must lay off the pepper! But I believe initially I didn't want the memory stick active, disc inserted switch to be reported by the prx and just left it as is. Should have just &= the values that i didn't want to be reported.

Soo humiliating but, sharing is caring....lol In any case its been fixed and updated and thank you for the constructive comments....lol

Ok... but i do have a question. I noticed when I do the following for analog mode:

Code: Select all

static int ctrl_button_thread&#40;SceSize args, void *argp&#41;
&#123;
	sceCtrlSetSamplingCycle&#40;0&#41;;
	sceCtrlSetSamplingMode&#40;PSP_CTRL_MODE_ANALOG&#41;;
	sceCtrlData paddata;
	while&#40;ctrl_active&#41;
	&#123;
		ctrl_data = 0;
		if&#40;pSceCtrlPeekBufferPositive&#41;
		&#123;
			&#40;*pSceCtrlPeekBufferPositive&#41;&#40;&paddata, 1&#41;;
			ctrl_data = paddata.Buttons;
			if&#40;paddata.Ly >= 0xC0&#41; ctrl_data |= PSP_CTRL_DOWN;
			if&#40;paddata.Ly <= 0x30&#41; ctrl_data |= PSP_CTRL_UP;
			if&#40;paddata.Lx <= 0x30&#41; ctrl_data |= PSP_CTRL_LEFT;
			if&#40;paddata.Lx >= 0xC0&#41; ctrl_data |= PSP_CTRL_RIGHT;
		&#125;
		sceKernelDelayThread&#40;10 * 1000&#41;;
	&#125;
	sceKernelExitDeleteThread&#40;0&#41;;
	return 0;
&#125;

The analog stick is not registered within the prx. But within the pbp it is, any ideas? I know its not a kernel mode issue versus user mode, but the behavior is quite different. Works in one and not the other.

hlide · Post by **hlide** » Sun Dec 30, 2007 8:10 am

SamuraiX wrote:I know, I know.... shame on me!!!!! Must lay off the pepper! But I believe initially I didn't want the memory stick active, disc inserted switch to be reported by the prx and just left it as is.

Soo humiliating but, sharing is caring....lol In any case its been fixed and updated and thank you for the constructive comments....lol

Ok... but i do have a question. I noticed when I do the following for analog mode:
Code: Select all
static int ctrl_button_thread&#40;SceSize args, void *argp&#41;
&#123;
	sceCtrlSetSamplingCycle&#40;0&#41;;
	sceCtrlSetSamplingMode&#40;PSP_CTRL_MODE_ANALOG&#41;;
	sceCtrlData paddata;
	while&#40;ctrl_active&#41;
	&#123;
		ctrl_data = 0;
		if&#40;pSceCtrlPeekBufferPositive&#41;
		&#123;
			&#40;*pSceCtrlPeekBufferPositive&#41;&#40;&paddata, 1&#41;;
			ctrl_data = paddata.Buttons;
			if&#40;paddata.Ly >= 0xC0&#41; ctrl_data |= PSP_CTRL_DOWN;
			if&#40;paddata.Ly <= 0x30&#41; ctrl_data |= PSP_CTRL_UP;
			if&#40;paddata.Lx <= 0x30&#41; ctrl_data |= PSP_CTRL_LEFT;
			if&#40;paddata.Lx >= 0xC0&#41; ctrl_data |= PSP_CTRL_RIGHT;
		&#125;
		sceKernelDelayThread&#40;10 * 1000&#41;;
	&#125;
	sceKernelExitDeleteThread&#40;0&#41;;
	return 0;
&#125;
The analog stick is not registered within the prx. But within the pbp it is, any ideas? I know its not a kernel mode issue versus user mode, but the behavior is quite different. Works in one and not the other.

Just some precisions :

sceCtrlSetSamplingCycle(0) --> set sampling at VSYNCH period

sceCtrlSetSamplingCycle(n) --> set sampling at n microsecond

sceKernelDelayThread(n) --> delay thread for n microsecond

it would probably be more efficient to do this way :

Code: Select all

static int ctrl_button_thread&#40;SceSize args, void *argp&#41;
&#123;
	// we want a sampling of 10 ms
	sceCtrlSetSamplingCycle&#40;10 * 1000&#41;;

	sceCtrlSetSamplingMode&#40;PSP_CTRL_MODE_ANALOG&#41;;
	sceCtrlData paddata;
	while &#40;ctrl_active&#41;
	&#123;
		ctrl_data = 0;
		if &#40;pSceCtrlReadBufferPositive&#41;
		&#123;                       
			// take the most recent state
			if  &#40;&#40;*pSceCtrlReadBufferPositive&#41;&#40;&paddata, 0&#41; > 0&#41;
			&#123;
				ctrl_data = paddata&#91;0&#93;.Buttons;
				...
			&#125;
		&#125;
	&#125;
	sceKernelExitDeleteThread&#40;0&#41;;
	return 0;
&#125;

this way, you will get a new state at each 10 ms. No need to sleep the thread, sceCtrlReadBufferPositive does it for you.

Also note, sceCtrlReadBufferPositive is no longer synched with VSYNCH event but has its own timer.

SamuraiX · Post by **SamuraiX** » Sun Dec 30, 2007 10:55 am

hlide wrote:this way, you will get a new state at each 10 ms. No need to sleep the thread, sceCtrlReadBufferPositive does it for you.

Also note, sceCtrlReadBufferPositive is no longer synched with VSYNCH event but has its own timer.

Hmm... without the sleep all I get is a blank screen when timer for ReadBuffer is set to 10 * 1000 (10 milliseconds). Also if I use ReadBuffer there is no control what so ever when sleep thread is used. So I think PeekBuffer is the only one that works in a kernel PRX.

But that still doesn't explain why Peek Buffer is not capable of reading the analog stick when running in a thread from a Kernel Prx.

hlide · Post by **hlide** » Sun Dec 30, 2007 2:17 pm

SamuraiX wrote:
hlide wrote:this way, you will get a new state at each 10 ms. No need to sleep the thread, sceCtrlReadBufferPositive does it for you.

Also note, sceCtrlReadBufferPositive is no longer synched with VSYNCH event but has its own timer.
Hmm... without the sleep all I get is a blank screen when timer for ReadBuffer is set to 10 * 1000 (10 milliseconds). Also if I use ReadBuffer there is no control what so ever when sleep thread is used. So I think PeekBuffer is the only one that works in a kernel PRX.

But that still doesn't explain why Peek Buffer is not capable of reading the analog stick when running in a thread from a Kernel Prx.

that's very weird... would Sony have messed with the kernel version in such a way we could not use them anymore except when in user mode ?

even with "u32 k1 = pspSdkSetK1(0); ...; pspSdkSetK1(k1);" around ? both for Read and Peek versions ?

SamuraiX · Post by **SamuraiX** » Sun Dec 30, 2007 3:53 pm

Just tried that too and no luck. It just doesn't make sense to only partially remove the functionality though.

forums.ps2dev.org

HOME button doesn't work on custom firmware... Why ?

HOME button doesn't work on custom firmware... Why ?

Re: HOME button doesn't work on custom firmware... Why ?

Re: HOME button doesn't work on custom firmware... Why ?

Re: HOME button doesn't work on custom firmware... Why ?

Re: HOME button doesn't work on custom firmware... Why ?