I'm looking at function stack sizes with the GCC -fstack-usage function.
In a sea of small stack sizes, I suddenly spot:

sysif.c:2067:12:crypt_wrap 131280 dynamic,bounded
^^^^^^

Jawdrop; 128K stack frame? Yagoddabekiddingme.

This is the current version, where I dropped the bullshit validation
code we discussed to death some weeks ago.

static val crypt_wrap(val wkey, val wsalt)
{ val self = lit("crypt");
const wchar_t *cwkey = c_str(wkey, self);
const wchar_t *cwsalt = c_str(wsalt, self);
char *key = utf8_dup_to(cwkey);
char *salt = utf8_dup_to(cwsalt);
#if HAVE_CRYPT_R
struct crypt_data cd;
char *hash = (cd.initialized = 0, crypt_r(key, salt, &cd));
#else
char *hash = crypt(key, salt);
#endif

free(key);
free(salt);

if (hash != 0)
return string_utf8(hash);

uw_ethrowf(error_s, lit("crypt failed: ~d/~s"), num(errno),
errno_to_str(errno), nao);
}

It has to be the crypt_data structure; and indeed it looks like this:

struct crypt_data
{
char keysched[16 * 8];
char sb0[32768];
char sb1[32768];
char sb2[32768];
char sb3[32768];
/* end-of-aligment-critical-data */
char crypt_3_buf[14];
char current_salt[2];
long int current_saltbits;
int direction, initialized;
};

Security through prodigality? :)

I would have made it sb[4][32768] if there were no way to evade the
extravagancy; but I suspect the author reveled also in the importune
obesity of the declaration.

--
TXR Programming Language: http://nongnu.org/txr
Cygnal: Cygwin Native Application Library: http://kylheku.com/cygnal