keyctl.2 updates

Hello.

While trying to add support for the new keyctl features (ability to provide KDF and ability to restrict addition to the keyring) decoding to strace, I've noticed that these are not documented, so here is my initial attempt to document these things. Unfortunately, keys/crypto infrastructure is quite foreign to me, so these patches definitely need some improvement.

For the one part is definitely omitted yet is the documentation for the "asymmetric" key type, which should be a part of keyrings.7, I think.

Thanks for all the feedback regarding the original version, I've tried to incorporate it in this version of the patchset.

Changes since v1:
* Added additional information about causes of EAGAIN, ENOENT, and ETIMEDOUT for the KEYCTL_DH_COMPUTE operation.
* Information about internal limit on hashname moved to the description of yet another cause of EINVAL for the KEYCTL_DH_COMPUTE.
* The description of possible restriction specification for asymmetric key type has been corrected.
* Added information about causes of ENOENT and EOPNOTSUPP for the KEYCTL_RESTRICT_KEYRING operation.
* Added wildcard mention of the ENOMEM error code.

Eugene Syromyatnikov (6):
keyctl.2: change "is" to "was" in error descriptions keyctl.2: update kernel documentation path reference keyrings.7: add pointers to kernel's documentation keyctl.2: document the ability to provide KDF parameters in
KEYCTL_DH_COMPUTE
keyctl.2: document KEYCTL_RESTRICT_KEYRING operation keyctl.2: mention ENOMEM in ERRORS

 man2/keyctl.2   | 289 +++++++++++++++++++++++++++++++++++++++++++++++++++++---
 man7/keyrings.7 |   5 +
 2 files changed, 282 insertions(+), 12 deletions(-)

Hi Eugene,

Quoted 8 lines · Eugene Syromyatnikov

Hello.

While trying to add support for the new keyctl features (ability to provide KDF and ability to restrict addition to the keyring) decoding to strace, I've noticed that these are not documented, so here is my initial attempt to document these things. Unfortunately, keys/crypto infrastructure is quite foreign to me, so these patches definitely need some improvement.

I've applied the all patches. If someone has further improvements, hopefully they send comments/fixes.

Quoted 2 lines

For the one part is definitely omitted yet is the documentation for the "asymmetric" key type, which should be a part of keyrings.7, I think.

That makes sense. Might you prepare a patch for that?

Quoted 23 lines

Thanks for all the feedback regarding the original version, I've tried to incorporate it in this version of the patchset.

Changes since v1:
* Added additional information about causes of EAGAIN, ENOENT, and ETIMEDOUT for the KEYCTL_DH_COMPUTE operation.
* Information about internal limit on hashname moved to the description of yet another cause of EINVAL for the KEYCTL_DH_COMPUTE.
* The description of possible restriction specification for asymmetric key type has been corrected.
* Added information about causes of ENOENT and EOPNOTSUPP for the KEYCTL_RESTRICT_KEYRING operation.
* Added wildcard mention of the ENOMEM error code.

Eugene Syromyatnikov (6):
keyctl.2: change "is" to "was" in error descriptions keyctl.2: update kernel documentation path reference keyrings.7: add pointers to kernel's documentation keyctl.2: document the ability to provide KDF parameters in
KEYCTL_DH_COMPUTE
keyctl.2: document KEYCTL_RESTRICT_KEYRING operation keyctl.2: mention ENOMEM in ERRORS

Quoted 3 lines

 man2/keyctl.2   | 289 +++++++++++++++++++++++++++++++++++++++++++++++++++++---
 man7/keyrings.7 |   5 +
 2 files changed, 282 insertions(+), 12 deletions(-)

So, I think everything that you sent has now been applied. Let me know if I mussed anything.