PriiLoader v0.5 Brick Warning

September 13th, 2010, 14:05 Posted By: wraggster

News via http://www.dacotaco.com/about-the-pr...er-0-5-removal

It's been a while since I removed the Priiloader 0.5 from my web server and the google code and I did not write about why I did (shame on me) .

I deleted the installer because of a bug. At the time of de - installation, there were a number of bricks ( and fortunately , in 99 % of cases , users had to BootMii boot2 > _ > ) because of this bug .

Why is this bug ?

easy

There was no Internet connection set. True , this is the reason the bug.

At first I was surprised how something like this can completely crash the installer ? ! Initially , I thought having a memory corruption . After a thorough cleaning of memory blocks of the installer , and other memory alignment , I could remove this idea from the list .

Maybe it was a bug in my code ? Only one way to find out . I started a basic "hello world " that initializes only the video and pad GC (so you can leave ) . And bam , crash . This led me to believe a problem with the HBC or libogc (made at this point, I was convinced that the HBC was concerned , because of AHBPROT now used to launch the application ) .

So we ( myself & lukegb ) continued testing and tested various simple operations dhewg help identify the problem. Time passed, and I was getting no response from him, so I tried the 2nd option : to inform the team of libogc this issue, hoping they would help me more.

And they did .

shagkur deleted this setting Internet and launched the demo application . No crash , he turned right. Oddly still, the basic " hello world " continued to grow , he then begin to look at the sources.

After a while he found the cause of the crash , caused by a response not expected CPI , which is connected with a callback registered (recorded from the CHB) . Where is the callback ? at 0x813a2028

This address can not say one thing: a response to a request ipc something pinnacle by the HBC , but never closed . Tantric & shagkur both believe that the source is the initialization of the asynchronous network from the HBC. They believe that this asynchronous request is never canceled / closed , and therefore the response ipc happens when the application is launched , causing this beautiful crash.

So yes , the crash is on the HBC and so it will not be corrected, I should not leave the other version of Priiloader This is not negotiable.

If we want to give the TT AHBPROT , they should check that everything is de- initialized properly ( without AHBPROT , HBC reload the IOS , losing the rights and re -initializing all from ipc ... So , no answer ipc unwanted in the application ) or that nothing bad could happen . That's what I spend most of my time working on the priiloader ( I have a lot of potential holes in my sources ; _ ; )

Geto Over and out

There are 0 comments - Join In and Discuss Here