radare

[john-peterson]

https://github.com/radareorg/radare2/issues?q=obr

Surely I can find some help

???

[0x180065508]> om
*12 fd: 3 +0x00000400 0x180001000 - 0x18006a9ff r-x fmap..text

[0x180065508]> obr 0x140000000
Segmentation fault (core dumped)          

(gdb) info file
                            Entry point: 0x140e53590
        0x0000000140001000 - 0x0000000140f719bc is .text

[john-peterson]

Would be nice if I could change the addresses

[john-peterson]

r2 -B 0x140000000 Metropolis_x64_vk.exe -B 0x180000000 FledgeCore_x64_vk.dll -B 0x002100000 FledgeFoundation_x64_vk.dll

NO??? ??????

[john-peterson]

o Metropolis_x64_vk.exe 0x140000000
o FledgeCore_x64_vk.dll 0x180000000
o FledgeFoundation_x64_vk.dll 0x002100000

It's seemingly or actually impossible to debug a program across multiple files

[john-peterson]

Also this is completely standard addressing. IT SHOULD DO THIS ITSELF look at those addresses. Do you think that's a coincidence

That's how msvc works!!!!!!!!

[john-peterson]

I SHOULDN'T HAVE TO EXPLAIN THR MOST BASIC FUNCTION OF MSVC ADDRESS ALLOCATION TO R2

[john-peterson]

Like all these apps it's just a russian disaster. Only way to fix us a separate fork. Pull requests would just start another Korean war. 1950 all over again. The Russian Revolution in Asia

[john-peterson]

Radiff

Absolute disaster from beginning to end

• stderr is rendered pointless by millions of repeated messages HOW CAN I SUPPRESS REPEAT MESSAGES?

• PRINT EVERY SINGLE FUNCTION AFTER IT IS DIFFED AND FREE MEMORY AND DO THE NEXT........ The current implementation is a disaster.... Should need six meg but need six gig

• Actually do your JOB! ADDRESS CHANGES ARE NOT CHANGES!!!!!!!!!!!!!!! it can't even... It's fundamentally flawed

Now it turns into tactfully fixing this. How can I put this in a polite manner and not turn everyone off. o:-) It is a miracle that it CAN find the corresponding functions. I am desperately looking for compliments here... I feel I already turned everybody off. By keeping talking

• The corresponding functions will be in an ADJACENT OFFSET no need to look through every single function. Would wipe 99% off the anal time

• Diffing itself has to EXCLUDE ALL, ALL, ADDRESS CHANGES even a single line of code will change EVERY address. This is the most major flaw because it's not just a a time waster for the anal job. It renders the entire output useless. 99% of UNMATCHED and NEW functions are in fact not that

...
....

🥞

....

[john-peterson]

I did one attempt to understand what it is and the pained maintainer pancakes... Well whatever. I don't want to talk about it

[john-peterson]

So we went from 600 000 lines to 62

Of actual changes

Just tell me

I. Mean

I have never seen anything that work better than r2. How can r2diff be a disaster

....

Pancakes

.....

[john-peterson]

It even has minor improvements over the mighty gdb

It will actually label msvc destructor constructor accurately and. Gdb is fine with just some random gibberish. That's enough for you!

[john-peterson]

They took pancakes from

It just works

To

If you can run this command without blowing up you win the hackathlon

[john-peterson]

sudo journalctl -u earlyoom | grep sending
Jul 30 03:27:52 deb earlyoom[12586]: sending SIGTERM when mem <= 10.00% and swap <= 10.00%,
Jul 30 03:35:38 deb earlyoom[12586]: sending SIGTERM to process 12490 uid 1001 "radiff2": badness 1214, VmRSS 3016 MiB
Jul 30 04:22:20 deb earlyoom[12586]: sending SIGTERM to process 16057 uid 1001 "radiff2": badness 1213, VmRSS 3011 MiB