scion+gmp@webrelay.net writes:
"ABI and ISA" in the manual describes the options you need to add to application compiles if you want to use the 64-bit ABI.
Indeed, That's the ticket. The link chain took me back to -lcrypto. Our libcrypto.so is a sparc-v8+ so that one pkg can live on all of our sparcs.
And so, compiling gmp, *and* nettle, (and pike and caudium) v8+ mode solved the problem.
Thanks to all who chimed in!
Cheers! -sam
From: Kevin Ryde user42@zip.com.au Date: Mon, 19 Jul 2004 08:36:28 +1000
...
./configure --disable-shared ABI=64
"ABI and ISA" in the manual describes the options you need to add to application compiles if you want to use the 64-bit ABI.
scion+gmp@webrelay.net writes:
Indeed, That's the ticket. The link chain took me back to -lcrypto. Our libcrypto.so is a sparc-v8+ so that one pkg can live on all of our sparcs.
And so, compiling gmp, *and* nettle, (and pike and caudium) v8+ mode solved the problem.
Note that this will make GMP run about 2x slower. Does v8 compatibility really matter, these days?
Ref: http://swox.com/gmp/32vs64.html
Probably not. We retired our farm of S1000s late last year. But all of the libraries that we use are still v8+. I'll need to recompile over half of our locally installed apps if I recompile libcrypto from OpenSSL. That lib makes it's presence known! Given that performance is not that big an issue (We're running Sparc/Solaris after all!), it'll take us at least the rest of the year to get it all out of the v8+ abi. By then, we may well be running Opteron/Solaris or even Opteron/FreeBSD.
-sam
From: Torbjorn Granlund tg@swox.com Date: 19 Jul 2004 17:26:17 +0200
scion+gmp@webrelay.net writes:
Indeed, That's the ticket. The link chain took me back to -lcrypto. Our libcrypto.so is a sparc-v8+ so that one pkg can live on all of our sparcs.
And so, compiling gmp, *and* nettle, (and pike and caudium) v8+ mode solved the problem.
Note that this will make GMP run about 2x slower. Does v8 compatibility really matter, these days?
Ref: http://swox.com/gmp/32vs64.html
nettle-bugs@lists.lysator.liu.se