Solar Designer

``` $ uname -mrsp SunOS 5.10 sun4u sparc ``` This says just `sparc`, not `sparc64`. So we'd have to check something like `uname -mrs` for regexp `^SunOS 5\.1. sun4u`. This...

Strictly speaking, jumbo's calling `sig_preinit()` and `path_init`() before CPU fallback is also a bug, until/unless we build `signals.c` and `path.c` without CPU type specific optimizations (like I think/hope we do...

> without CPU type specific optimizations (like I think/hope we do for john.c, or don't we?) We do (or at least can) in core, due to these lines in `Makefile`:...

> jumbo's calling `sig_preinit()` and `path_init`() before CPU fallback is also a bug, until/unless we build `signals.c` and `path.c` without CPU type specific optimizations We no longer call `sig_preinit` before...

See @claudioandre-br's work in #2277 and #2280. The problem is in jumbo we sometimes use `$JOHN` for the fallback path: ```c #ifdef JOHN_SYSTEMWIDE_EXEC #define CPU_FALLBACK_PATHNAME JOHN_SYSTEMWIDE_EXEC "/" CPU_FALLBACK_BINARY #else #define...

> Do You (the project) has any update on this? Unfortunately, no. I'm very familiar with our bitslice DES CPU code (having been the one to introduce it and update...

> I disabled mscash2 for another reason (which I will find out sooner or later). It works on a BE system. We had #5190, which I fixed, however it should...

Solaris 10, sparc64, 32-bit `configure` build, OpenMP: ``` Testing: timeroast, SNTP-MS [MD4+MD5 32/32]... (3xOMP) FAILED (cmp_all(1)) ``` Solaris 11, sparc64, 64-bit `Makefile.legacy` non-OpenMP build: ``` Testing: timeroast, SNTP-MS [MD4+MD5 32/64]......

I figured this one out, it will get closed by an upcoming PR shortly.

> it's very hard to spot where a standard SHA-2 operation starts or ends I guess it would be easier to rewrite based on the CPU format, which is pretty...