6 ms·

if ((err = SSLHashSHA1.update(&hashCtx, &signedParams)) != 0) goto fail; goto fail; /* MISTAKE! THIS LINE SHOULD NOT BE HERE */ if ((err = SSLHas

by IncRnd 9y ago

if ((err = SSLHashSHA1.update(&hashCtx, &signedParams)) != 0) goto fail; goto fail; /* MISTAKE! THIS LINE SHOULD NOT BE HERE */ if ((err = SSLHashSHA1.final(&hashCtx, &hashOut)) != 0) goto fail; History repeats itself in ever stranger ways. While the effects are different, the original mistake can be quite similar.

saagarjha 9y ago
This is the SSL bug from iOS 9 for those wondering.
- caseymarquis 9y ago
  Neat that the textbook justification for Python whitespace syntax over curly brace syntax has caused serious real world bugs.
emsy 9y ago
You don't even need static analysis (which they should use too) but just a strict coding style check that enforces curly braces. Should've happened after the first incident!
- tempodox 9y ago
  I find it jaw-dropping how fashionable cowboy coding still is, especially in C family languages.
  - eru 9y ago
    C-family languages are cowboy languages. The whole point of eg C (and C++)'s undefined behaviour is to allow the compiler to make cowboy assumptions like "this array access will never be out of bounds" or "this signed int will never overflow" without having to prove or even justify them. All in the name of 'efficiency'.
- eru 9y ago
  A coding style check _is_ static analysis. (But that's beating a straw man, of course: with that expanded definition your comment just because "you don't even need anything more than the most primitive forms of static analysis".)
  - emsy 9y ago
    I think the point I was trying to make was: If they don't even check for rudimentary things like coding style, can we expect them to check for more sophisticated sources of issues (e.g. memory corruption).
    - eru 9y ago
      Amen.
- astrange 9y ago
  The Python and Haskell programmers would be upset if you added that pre-commit hook.
devdoomari 9y ago
um... use dependent types? sorry but I'm from different language background :(
- IncRnd 9y ago
  Provable code would solve the root cause, yes. :) However, even with provable code, the proof actually has to be both correct and performed. There are too many competing factors that won't allow idris, for example, to actually work. The result is that, as an industry, the Internet and the world's business are held together by twine, twist ties, and spaghetti code. Even this very webpage is just enough to work for most people a lot of the time.
  - devdoomari 9y ago
    so... I guess strong-AI-based coding is the only 100% sure way to go (but... I'd be jobless if that were to happen :(
    - eru 9y ago
      Strong AI wouldn't be enough, if by strong AI you mean "enough to pass the Turing test". Humans pass the Turing test all the time, but still produce buggy software.
    - IncRnd 9y ago
      There are some insanely good architects and programmers. Even so, people write mistaken code, design with faulty assumptions, misinterpret business logic, use libraries and crypto incorrectly, etc. On top of all that, there is a small segment of coders who insert malicious code. They used to write for mayhem but now more for profit. The terms are somewhat contentious, but AIs are more prone to use heuristics than traditional algorithms, which would add another layer of complexity. So, we can't be certain that such an AI itself wouldn't create bugs. If anything, it would be easier to show that bugs would get created. This is simply the state of software. We all deal with it in the ways we can, trying our best to minimize issues and add value.
- 9y ago

by IncRnd 9y ago

saagarjha 9y ago

This is the SSL bug from iOS 9 for those wondering.

caseymarquis 9y ago
Neat that the textbook justification for Python whitespace syntax over curly brace syntax has caused serious real world bugs.

emsy 9y ago

You don't even need static analysis (which they should use too) but just a strict coding style check that enforces curly braces. Should've happened after the first incident!

tempodox 9y ago
I find it jaw-dropping how fashionable cowboy coding still is, especially in C family languages.
- eru 9y ago
  C-family languages are cowboy languages. The whole point of eg C (and C++)'s undefined behaviour is to allow the compiler to make cowboy assumptions like "this array access will never be out of bounds" or "this signed int will never overflow" without having to prove or even justify them. All in the name of 'efficiency'.
eru 9y ago
A coding style check _is_ static analysis. (But that's beating a straw man, of course: with that expanded definition your comment just because "you don't even need anything more than the most primitive forms of static analysis".)
- emsy 9y ago
  I think the point I was trying to make was: If they don't even check for rudimentary things like coding style, can we expect them to check for more sophisticated sources of issues (e.g. memory corruption).
  - eru 9y ago
    Amen.
astrange 9y ago
The Python and Haskell programmers would be upset if you added that pre-commit hook.

devdoomari 9y ago

um... use dependent types? sorry but I'm from different language background :(

IncRnd 9y ago
Provable code would solve the root cause, yes. :) However, even with provable code, the proof actually has to be both correct and performed. There are too many competing factors that won't allow idris, for example, to actually work. The result is that, as an industry, the Internet and the world's business are held together by twine, twist ties, and spaghetti code. Even this very webpage is just enough to work for most people a lot of the time.
- devdoomari 9y ago
  so... I guess strong-AI-based coding is the only 100% sure way to go (but... I'd be jobless if that were to happen :(
  - eru 9y ago
    Strong AI wouldn't be enough, if by strong AI you mean "enough to pass the Turing test". Humans pass the Turing test all the time, but still produce buggy software.
  - IncRnd 9y ago
    There are some insanely good architects and programmers. Even so, people write mistaken code, design with faulty assumptions, misinterpret business logic, use libraries and crypto incorrectly, etc. On top of all that, there is a small segment of coders who insert malicious code. They used to write for mayhem but now more for profit. The terms are somewhat contentious, but AIs are more prone to use heuristics than traditional algorithms, which would add another layer of complexity. So, we can't be certain that such an AI itself wouldn't create bugs. If anything, it would be easier to show that bugs would get created. This is simply the state of software. We all deal with it in the ways we can, trying our best to minimize issues and add value.
9y ago