let's test improving menuhax boot rate (ninjhax 2.5)

so, I've been modifying the ninjhax 2.5 source code, removing services to try and improve boot rate. this is not something I came up with, I saw @Mrrraou tweet about it at first, and I wanted to have a go at it.

this removes services that ninjhax 2.5 tries to get for homebrew to use. it has been narrowed down to the only four that are required for ninjhax, a boot manager, and custom firmware (or things like Decrypt9 since they run their own exploits) to run.

unfortunately I can't easily test boot rate quickly and for other regions, which is why this post exists. when using these ropbins for menuhax, do you see an increase or decrease in boot rate for menuhax? try using it in normal use and see.

these do not remove debug colors or text at the moment.

thanks in advance!

https://github.com/ihaveamac/ninjhax2.5-less_services_test/releases

check out my changes: https://github.com/ihaveamac/ninjhax2.5-less_services_test/commits/master
  • Like
Reactions: 7 people

Comments

EpicLPer
Actually my Menuhax bootrate is amazing so far, only 1 out of 20 tries fail on both my New and Old 3DS's.
 
D
Worth a try, since the official MenuHax 2.1/2.2 payloads downloaded result in a terrible boot rate on my N3DS.
 
D
Doesn't seem to do much for the infamous white screen problem, unfortunately (which is to be expected, I suppose, since that's menuhax failing, not the payload). Roughly 50% success rate, if that, but the hax payload itself is booting Reinand every other time. Curse you menuhax..
 
ihaveahax
"since that's menuhax failing"

really? I think if the payload is missing or corrupt, it will not change to a white screen at all...
 
D
But it's not missing or corrupt. It's just..menuhax failing somewhere, hanging on a white screen.
 
ihaveahax
I could have worded that better, but I mean once it changes to a white screen I thought it's the payload failing if it doesn't change. but it also only happens with menuhax so it could very well be that
 
endoverend
What's up with the white screen thing? it like slowly fades into white with some barely visible text behind it and it's actually scary
 
ihaveahax
don't know... it only happens on my New3DS. it just freezes at white sometimes. similar to the yellow screen freeze but only with menuhax.
 
D
"I could have worded that better, but I mean once it changes to a white screen I thought it's the payload failing if it doesn't change. but it also only happens with menuhax so it could very well be that"

Oh, I'm not sure then, to be honest. It'd be awesome if we could figure out where the problem lies.
 
EpicLPer
So you tell me there's a white screen freeze?
I never had that happening on my EU N3DS or O3DS... Max. I got was a yellow one.
 
D
It's a problem for a number of people (see the menuhax issue tracker), and there's no known solution aside from using an old payload from menuhax 1.3 (aka themehax). It's annoying.
 
D
Come to think of it, maybe it IS the payload failing, since when i go back to the ancient themehax 1.3 payload (so hax 2.x from the end of September 2015), it never gets stuck on a white screen, even if I'm using menuhax 2.1/2.2 (yes, this is possible).
 
Mrrraou
Works fine for me. 19/20 success rate. Only one yellow screen. O3DS 9.0.0-20E.
I guess that the sleep time on blue screen can be reduced, tho.
 
Februarysn0w
oh....it's worked. I had been something wrong yesterday. boot rate is so stable. reinand always show me boot image. very cool. thank you for your nice release.
 
G
Hey. Shot you an issue over on GitHub. Just found this thread here.

Why is ndsp necessary? It's just sound, and it most definitely can be edited out as older version of the payload didn't even have it.
Won't removing AM:Sys break some hacktools that launch from HomeBrew launcher?
MEM0 seems unnecessary as well if our core goal is to boot into a CFW and allow some HBL hack tools.

We should be able to narrow it down to simply:

- fs:USER for obvious reasons.
- ns:s for obvious reasons.
- am:sys to allow some hack tools to still launch from HBL.

Which is the original 4 minus the IR service (why was that even there???).

If am:sys isn't necessary for ANY hack tool, we could simply narrow it down to 2.

Also, I'm suggesting a motion to rebrand this build of NinjHax without debug info "LessHax" :D
 
  • Like
Reactions: 1 person
Mrrraou
@gameking66 I guess you're right on that. However, how can we be sure that hb:mem0 isn't used ? Is there any info about what is it can be used for precisely ? (I checked the hb.c in ctrulib repo.) I guess it won't break anything, anyway.
 

Blog entry information

Author
ihaveahax
Views
268
Comments
120
Last update

More entries in Personal Blogs

More entries from ihaveahax

Share this entry

General chit-chat
Help Users
  • No one is chatting at the moment.
    NinStar @ NinStar: CRAZY HAMBURGER