Thanks for reporting this. It looks like a bug in this old version of gas, 
but I'll add a workaround for it.

--------------------------------------------------
From: "Paul de la Iglesia" <[email protected]>
Sent: Friday, February 20, 2009 5:16 PM
To: "Crypto++ Users" <[email protected]>
Subject: Error in CryptoPP 5.2.2 when compiling sosemanuk.cpp on 
x86_64-suse-linux / gcc v3.4.2 / gas v2.15

>
> Here's my setup
>
> g++/gcc version 3.4.2
> host x86_64-suse-linux
> gas version 2.15 using BFD version 2.15
>
> Here's my error
>
> # make -f GNUmakefile
> g++ -DNDEBUG -g -02 -pipe -c 3way.cpp
> ... *SNIP* good compiling mojo here ...
> g++ -DNDEBUG -g -02 -pipe -c sosemanuk.cpp
> {standard input}: Assembler messages:
> {standard input}:1046: Error: `[r9d+edx]' is not a valid 64 bit base/
> index expression
> {standard input}:1046: Error: `[r9d+ecx]' is not a valid 64 bit base/
> index expression
> ... *SNIP* repeat previous 2 lines 8x ...
> {standard input}:1046: Error: `[r9d+edx]' is not a valid 64 bit base/
> index expression
> {standard input}:1046: Error: `[r9d+ecx]' is not a valid 64 bit base/
> index expression
> make: *** [sosemanuk.o] Error 1
> #
>
> Adding -DCRYPTOPP_DISABLE_ASM to g++ eliminates the error shown
> below.  The "cryptest.exe v" tests all pass, so it works as far as I
> can see.  However, I'm still very new to g++ and the assembler, so I
> don't know if I'm masking some bad behavior with that flag.
>
> I'd appreciate any feedback on whether using that flag will bite me
> later on.  If this is an acceptable fix, then yay to me and thanks for
> the feedback.  I only saw this error in one other post in this group,
> and I didn't see a resolution to that one, so I hope this helps other
> people that may run accross the issue.
>
> --Paul
> > 

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the "Crypto++ Users" 
Google Group.
To unsubscribe, send an email to [email protected].
More information about Crypto++ and this group is available at 
http://www.cryptopp.com.
-~----------~----~----~----~------~----~------~--~---

Reply via email to