Ok I've got you attention . I'll play the devils advocate a bit here. What if someone write (maliciously or not) a HEX that would use SPM instructions in an endless loop all over the flash memory? It's clear that at 10K max rewrites, you are gonna kill your chip in no time. Since I'm an incurable internet paranoid, I'd suggest you protect yourself against that and set the protection fuse on you mega644! Eventually we will have a SD bootloader and it should be (with the ISP) the only way to 'wear' your chip.
Uze
SPM Trick: Kill your Uzebox in 1 sec!
Re: SPM Trick: Kill your Uzebox in 1 sec!
Its time for Norton Antivirus Uzebox Edition
Re: SPM Trick: Kill your Uzebox in 1 sec!
I understand your concern Uze, but this argument comes up at work a lot for me. We do datalogging to EEPROM on our embedded boards and it inevitable comes out that "Oh No....we only have 10-20k write cycles on the EEPROM. Won't it wear out?" And yes, its a possibility. But I mean, if someone writes a malicious program for the Uzebox, there is only one filter to stop it: users.
Abusing the hardware is just something you can't really protect against. And I mean, if worst comes to worst, you have to buy another 5-8$ chip and snap it on your uzebox. I say its not that big a deal. I have a backup 644 just in case I need it anyways. Its not like you have to throw the whole board away.
And besides...who here would write a bad program like that anyways? ...at least not intentionally, right?
Abusing the hardware is just something you can't really protect against. And I mean, if worst comes to worst, you have to buy another 5-8$ chip and snap it on your uzebox. I say its not that big a deal. I have a backup 644 just in case I need it anyways. Its not like you have to throw the whole board away.
And besides...who here would write a bad program like that anyways? ...at least not intentionally, right?
Re: SPM Trick: Kill your Uzebox in 1 sec!
You guessed my secret game. Congratsuze6666 wrote:What if someone write (maliciously or not) a HEX that would use SPM instructions in an endless loop all over the flash memory?
>J
Lerc wrote:I intend to use my powerful skills of procrastination to ensure that when I get to making things, the chips will be available.
Re: SPM Trick: Kill your Uzebox in 1 sec!
True...anyway, I just had to mention it. Kind of a disclaimer...Its not like you have to throw the whole board away.
Uh oh, the first banned user! ...You guessed my secret game. Congrats
Uze
-
- Posts: 252
- Joined: Tue Dec 02, 2008 12:38 am
- Location: Carlsbad, California (USA)
Re: SPM Trick: Kill your Uzebox in 1 sec!
Another option would be to run an "untrusted" hex under the emulator first. It currently warns about EEPROM access but not SPM just yet.uze6666 wrote:Ok I've got you attention . I'll play the devils advocate a bit here. What if someone write (maliciously or not) a HEX that would use SPM instructions in an endless loop all over the flash memory? It's clear that at 10K max rewrites, you are gonna kill your chip in no time. Since I'm an incurable internet paranoid, I'd suggest you protect yourself against that and set the protection fuse on you mega644! Eventually we will have a SD bootloader and it should be (with the ISP) the only way to 'wear' your chip.
Uze
-Dave
Re: SPM Trick: Kill your Uzebox in 1 sec!
I'll have to work vigilantly to release it before the next emulator release now that I know this.Another option would be to run an "untrusted" hex under the emulator first. It currently warns about EEPROM access but not SPM just yet.
Anyway in all seriousness, is there a way to accidentally do something like this in C, or just if your doing assembly?
Is there any chance that if something runs on the emulator it could crash the actual hardware?
Lerc wrote:I intend to use my powerful skills of procrastination to ensure that when I get to making things, the chips will be available.
-
- Posts: 252
- Joined: Tue Dec 02, 2008 12:38 am
- Location: Carlsbad, California (USA)
Re: SPM Trick: Kill your Uzebox in 1 sec!
Definitely, because I emulate only a small percentage of the real hardware on an AVR8, just enough to get working input, video, and sound.Jhysaun wrote:I'll have to work vigilantly to release it before the next emulator release now that I know this.Another option would be to run an "untrusted" hex under the emulator first. It currently warns about EEPROM access but not SPM just yet.
Anyway in all seriousness, is there a way to accidentally do something like this in C, or just if your doing assembly?
Is there any chance that if something runs on the emulator it could crash the actual hardware?
But I would like to think we're a small enough group with narrow enough interests that we'd be free of any malice.
-Dave