--
--
You received this message because you are subscribed to the "Crypto++ Users" Google Group.
To unsubscribe, send an email to cryptopp-user...@googlegroups.com.
More information about Crypto++ and this group is available at http://www.cryptopp.com.
---
You received this message because you are subscribed to the Google Groups "Crypto++ Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cryptopp-user...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
I'm fine with this as long as it doesn't change behavior on Windows / Visual Studio, which normally catches asserts and uncatched exceptions by default :)
I'm catching an assert when running with -DDEBUG (i.e., not -DNDEBUG):
cryptest.exe: misc.h:703: T CryptoPP::rotlMod(T, unsigned int)
[with T = unsigned int]: Assertion `y <= 255' failed.
The assert is fine. I adore asserts, because they create "self debugging code". When the code tells me where the problem is, I don't have to spend time under a debugger. In code under my purview in real life, we make debug instrumentation a security gate because it so helpful. If code does not have the instrumentation, it cannot be checked in.
The issue I have is Posix's default behavior of "lets crash the program while the developer is debugging it". Its completely useless behavior.
Attached is the meat and potatoes of the proposed patch. It omits all the assert -> CRYPTOPP_ASSERT changes.
It works great outside and under the debugger. The program does not crash with a SIGABRT at critical times during the debug cycle.
You may want to add that #define CRYPTOPP_INSTALL_SIGTRAP_HANDLER (in the comment) only affects GCC / Unix.
I'm not sure if everybody knows this is unix-only behavior.
I use -Wall, GCC 4.8.1
By the way, there are lots of other warnings, not with asserts.
I can't say for sure right now, but there were unused StringNarrow method and unused *p pointer. I'll send more details tomorrow.
Well... Actually, no; I use Crypto++ 5.6.2 from cryptopp.com. Am I wrong?
OK, I'll try to recompile it either this evening or tomorrow.
The problem I've reported on the 26th of July is still actual.