Re: {twofish,aes}-{x86_64,i586} versus C implementations
From: Sebastian Siewior <hidden>
Date: 2007-10-04 09:39:28
* Herbert Xu | 2007-10-04 16:48:18 [+0800]:
On Thu, Oct 04, 2007 at 10:35:12AM +0200, Sebastian Siewior wrote:quoted
Two last questions: - What about the i386 assembly vs generic implementation? Do you prefer the patch that I have send earlier (choose the assembly by default making the generic optional) or do you want both of them loaded at the same time.I'd prefer both to be built by default so that if something does go wrong we can ask people to check by using aes-generic.
That sounds fair to me. I don't touch twofish since it is not auto selected like aes is and the assembly version is available by default unless the user enabled both. Maybe a hint in Kconfig's help would help but I don't thing it is necessary. The distros should not compile it in first place :)
quoted
- What might be the best fallback strategy for the s390 + geode aes driver?I've been trying to avoid this :) Yes your proposal is definitely on the right track. My original plan is to have a template like keylen(aes,16) that only supports key length 16. So perhaps you can take a look at doing that or come up with a different solution.
Okey. So I drop my changes to the driver and try to fix the crypto API like I was doing in first place.
Cheers,
Sebastian