1

Topic: JIT compile

Prompt please, whether as in application to define the method in this application and if it was compiled how to learn the address of the beginning of the machine code of this method was compiled? Ps: read 2 articles about a code injection in runtime. One with source codes and did not make the elementary injection, and after testing of the second method the computer started to show a porno the goods though the injection assumed addition changeover on ))

2

Re: JIT compile

Hello, SanyaVB, you wrote: SVB> Prompt please, whether as in application to define the method in this application and if it was compiled how to learn the address of the beginning of the machine code of this method was compiled? Depends on the task. It is possible to be perverted / CLR MD, it is possible to get into guts with dirty hands, the correct method - not to do generally.

3

Re: JIT compile

Hello, Sinix, you wrote: S> Depends on the task. It is possible to be perverted / CLR MD, it is possible to get into guts with dirty hands, the correct method - not to do generally. The first article not so... It one of those about which I wrote in the beginning. A lot of indirect With ++ the code, and  does not correspond to a source code (if someone to dare to compile all the same a source code - here so simply it does not turn out, is easier Linux hands to collect)... More shortly . And here the second link was pleasant to me. This that that is necessary for a campaign! Thanks!

4

Re: JIT compile

Hello, SanyaVB, you wrote: SVB> And here the second link was pleasant to me. This that that is necessary for a campaign! Thanks! With probability in 99 % is not that it is necessary for you. Long story short to climb in guts CLR and to change values directly approximately as safely and reliably, as   bit in address space of another's process. Decisions are, but all of them not the general-purpose. It is necessary to know the exact task to prompt more precisely.

5

Re: JIT compile

Hello, SanyaVB, you wrote: SVB> Prompt please, whether as in application to define the method in this application and if it was compiled how to learn the address of the beginning of the machine code of this method was compiled? SVB> Ps: read 2 articles about a code injection in runtime. One with source codes and did not make the elementary injection, and after testing of the second method the computer started to show a porno the goods though the injection assumed addition changeover on )) Not this? Interception of functions.NET/CLR

6

Re: JIT compile

Hello, Sinix, you wrote: S> Decisions are, but all of them not the general-purpose. It is necessary to know the exact task to prompt more precisely. Software protection. The idea consists in that that certain methods were in a useless type (i.e. do not carry in itself the useful operation), but in runtime they can change the functional and after performance to return itself in an initial state (as though to clean traces in storage). Will change a functional a separate flow (manage or native has no value)

7

Re: JIT compile

Hello, SanyaVB, you wrote: S>> Decisions are, but all of them not the general-purpose. It is necessary to know the exact task to prompt more precisely. SVB> software protection. The idea consists in that that certain methods were in a useless type... In  to pull out easier. All remaining methods or manage through standard API, or spoil application by a various and unpredictable method. Most  from this that came across - application-suitsidnik: through any time after setting fulfilled Auto NGen Maintenance Task   Keywords for search: NGen indirections + NGen fix-up tables. If it is short, that though approximately to understand, how to force this business to work, it is necessary to shovel mountain . And 99.9 that any other method except advances with JIT "will be the total decision". UPD: And, well or replace direct calls with calls of delegates (the main thing not to store the delegate in static a readonly-field), a body of the delegate fill dynamic. On productivity  will be, and adventures pick up, but it though will work. In the core.

8

Re: JIT compile

Hello, Sinix, you wrote: S> UPD: And, well or replace direct calls with calls of delegates (the main thing not to store the delegate in static a readonly-field), a body of the delegate fill dynamic. On productivity  will be, and adventures pick up, but it though will work. In the core. Well... Yes! Will work. But any beginner can crack such protection. And then there is a question - whether it is necessary to pile up these delegates or to be restricted to a simple condition? if (trial) {} else {} on speed of breaking and kol-vu which implementation and debugging - bright can crack people approximately equally, but

9

Re: JIT compile

Hello, SanyaVB, you wrote: SVB> Well... Yes! Will work. But any beginner can crack such protection. Well complicate it the task - collect a code part dynamic. Or, say, make the dictionary the guid-delegate, fill it on 99 % with garbage, 1 % - the correct methods created through expression tree. The table to fill in , the correct keys to methods to substitute in . The sense in it is not present any, but 99 % of beginners frightens off. SVB> and then there is a question - whether it is necessary to pile up these delegates or to be restricted to a simple condition? It is not necessary. The majority of products generally reduce check to a key in the register or to a file in %appdata %. And , live somehow. People which  to spend $10-20 for the useful tool hardly generally are your clients, the sense with them to struggle? Make unpretentious check - store the installation date signed by a private key and change a key each release. The key is not present - ask to reinstall. All.