src - OpenBSD base system

Age	Commit message (Collapse)	Author
2002-11-21	From Angelos:	Jason Wright
	- simplistic load balancing across multiple cards - simplified registration process - a few style nits.
2002-11-12	Add a CRYPTO_NULL xform (it's a do nothing, but nice for measuring the	Jason Wright
	bandwidth of the kernel API). It's only available from userland and then only if kern.cryptodevallowsoft=1.
2002-10-27	Use int32_t rather long on strategic places, to be more sparc64-friendly;	Miod Vallat
	ok jason@ a long time ago
2002-08-16	fix size in MALLOC/bzero call, from sam@errno.com, ok deraadt@	Daniel Hartmeier

2002-07-17	I don't know why this breaks things for me when sshd starts on sparc64.	Artur Grabowski
	But after wasting the whole day trying to just locate the problem I don't care. Back out since this wasn't tested and showed to anyone else.
2002-07-16	Double-pass over drivers, first hardware only, then software (if we	Angelos D. Keromytis
	are interested in software).
2002-07-16	Fix a typo, cleanup on session migration code in crypto_invoke(), and	Angelos D. Keromytis
	add a convention that if the driver returns ERESTART as an error message of its process method, the crypto framework will unregister the driver and migrate all its sessions. After discussion with Sam Leffler and Jason Wright.
2002-07-16	Add CRYPTO_ALGORITHM_ALL	Angelos D. Keromytis

2002-07-10	proper cdev_decl(crypto) and no private protos	Michael Shalayeff

2002-07-10	use right sizeof in memcpy; markus ok	Theo de Raadt

2002-07-02	inital -> initial	Nathan Binkert

2002-06-19	ansi	Theo de Raadt

2002-06-19	re-enable both crypto operations	Theo de Raadt

2002-06-12	looks slightly better	Theo de Raadt

2002-06-12	undo 1.46; fixes double free (the keys are freed on session close).	Markus Friedl
	passes regress/sys/crypto
2002-06-12	so let us analyse the development process here:	Theo de Raadt
	- complete breakage of symmetric userland crypto - fast commit without review by other developers - no regression test run (markus points out regression crash kernel) Hence, we must disable userland crypto for now, until it gets repaired by these people who don't follow process....
2002-06-11	kernel changes to make asymmetric crypto work in userland	Bob Beck
	- modify getfeat to return something more useful to us on devices (like lofn and everything else until jason fixes it) that can't do rsa stuff, etc and can only do mod_exp.. - error handling fixes so we correctly fail to software when we can't deal with a particular key size - add sysctl kern.userasymcrypto to turn on/off userland asymmetric crypto via /dev/crypto - 1 == on, 0 == off, default is off
2002-06-10	fix ivory tower greek fix. ok angelos@	Bob Beck

2002-06-10	You should always free the keys, not just on error --- the crypto	Angelos D. Keromytis
	framework doesn't hold on to these.
2002-06-10	__FUNCTION__ -> __func__	Marc Espie

2002-06-09	Don't use an int for the flags, when the structure uses	Angelos D. Keromytis
	u_int8_t. Also, make sure the logic is correct (bad theo!)
2002-06-07	Handle csecreate errors. deraadt@ and beck@	Thomas Nordin

2002-05-18	not ready to export fcrypt to userland yet; spotted by lebel	Theo de Raadt

2002-05-18	export struct fcrypt to userland	Theo de Raadt

2002-05-08	Don't deref null pointer in failure case.	Jason Wright

2002-04-27	32 bit constraint is wrong, we think	Theo de Raadt

2002-04-26	minor tweaks	Theo de Raadt

2002-04-26	disable cryptodevallowsoft again; constrain key bit sizes to % 32cryptodev.c	Theo de Raadt

2002-04-24	kernel API does not need the userland addresses, make them go away	Theo de Raadt

2002-04-23	initial hack at a CIOCSYMFEAT ioctl	Theo de Raadt

2002-04-23	driver queueing & callback code for keying operations	Theo de Raadt

2002-04-22	KNF	Theo de Raadt

2002-04-08	Credit DARPA/USAF appropriately.	Jason Wright

2002-04-03	fix possible alignment problem; with markus@, angelos@ deraadt@ ok.	Federico G. Schwindt

2002-04-03	disable userland crypto for 3.1	Theo de Raadt

2002-03-19	Don't keep the last blocksize-bytes of ciphertext for use as the next	Angelos D. Keromytis
	plaintext's IV, in CBC mode. Use arc4random() to acquire fresh IVs per message instead (particularly useful for IPsec). This avoids the CBC oracle attack. provos@ ok
2002-03-15	Kill #if __STDC__ used to do K&R vs. ANSI varargs/stdarg; just do things	Todd C. Miller
	the ANSI way.
2002-03-14	First round of __P removal in sys	Todd C. Miller

2002-03-05	export MD5/SHA1 via /dev/crypto; ok provos@, beck@	Markus Friedl
	tested with cryptosoft and kern.cryptodevallowsoft=1
2002-03-04	parts that copy in PK parameters	Theo de Raadt

2002-03-04	remove old PK stuff	Theo de Raadt

2002-03-04	crypto_check_alg() is not needed	Theo de Raadt

2002-03-02	do some more renaming	Theo de Raadt

2002-03-02	initial cut at crypto_sop definitions for symmetric crypto	Theo de Raadt

2002-03-01	sysctl for cryptodevallowsoft - /dev/crypto will allow the software engine	Niels Provos
	to be used if there are no hardware cards. mostly for debugging and regression.
2002-03-01	remove CRYPTO_BUF_CONTIG and convert to handle iovs. okay deraadt@	Niels Provos

2002-03-01	helper functions for uio (cuio_apply and cuio_getptr) okay deraadt@	Niels Provos

2002-02-24	license repair, angelos ok	Theo de Raadt

2002-02-23	sysctl kern.usercrypto	Theo de Raadt

2002-02-08	- Rename FILE_{,UN}USE to FREF and FRELE. USE is a bad verb and we don't have	Artur Grabowski
	the same semantics as NetBSD anyway, so it's good to avoid name collissions. - Always fdremove before freeing the file, not the other way around. - falloc FREFs the file. - have FILE_SET_MATURE FRELE the file (It feels like a good ortogonality to falloc FREFing the file). - Use closef as much as possible instead of ffree in error paths of falloc:ing functions. closef is much more careful with the fd and can deal with the fd being forcibly closed by dup2. Also try to avoid manually calling *fo_close when closef can do that for us (this makes some error paths mroe complicated (sys_socketpair and sys_pipe), but others become simpler (sys_open)).