I have been using VC with my new HP Spectre for a few months now and have been experiencing some strange behavior I would like to ask you about (system details see below).
I use a YubiKey with a fixed password plus a 4-digit pin. So the pre-boot sequence is (pin) + (yubikeypass) + (enter, from yubikey). No Pim.
It started to receive once in a while a “failed password, hash or pim” message on the first try only. Now I receive such a fail on first attempt on all boot processes. However, most of the time the second attempt works. Strangely the second attempt-fails are growing…. then third attempt works. So it seems I have “evolved” from 0 to 100% first attempt fails and now moving towards 100% 2nd attempt fails… (???)
Anyone an Idea what this is about and how I can fix it? So far I could always enter Windows, it is just annoying for the moment.
Thank you!
Michael
System:
HP Spectre x360 Convertible 13-ap0xxx, BIOS F.07
EFI Version 2.70, x64
VeraCrypt 1.23 (x64)
I've got a similar setup: HP Spectre 360 running Win 10 with Veracrypt v1.23 and using YubiKey to enter the pre-boot password.
My issue is that regardless of the password length (tried 64 and 32 char) I program into the key, VC only accepts the first 16-20 characters before moving on to the PIM (still not sure what that is) and the authorization fails. I can still enter the (very long, very complex) password manually for access.
The Yubikey dumps the complete password into a text editor just fine, so I'm prety sure its programmed correctly. Anyone know a reason why VC won't accept the entire password?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I'm having the same issue with a machine with EFI enabled. Despite using the YubiKey Personalization Tool to set a delay of 60ms, I'm experiencing the same issue as @msm1111 above. Once you're booted into Windows, everything is fine so it must be the EFI boot process that changed recently with VC 1.23.
I'm currently using VC 1.23 Hotfix 2 and from what I recall, I didn't have this issue pre VC 1.23. @idrassi, would it be possible to add a configurable key delay at the bootloader?
Thank you!
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Its been couple of months. And i would like to refresh this topic, as it is still valid request. Can we get a confirmation that it will be soved one day or should we start looking for another soluton?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
@msm1111, @jhaessly, @alancode, @fastriffs: sorry for the late answer. VeraCrypt EFI bootloader has a 100ms delay between accepted key strokes and so if a key received less than 100 ms before previous key then it will be ignored. This mechanism was added to fix issues with some keyboards that send keystores in double during a short period.
Unfortunately, it looks like yubikey has a maximum delay of 60ms between keystrokes, so most certainly some of the keystrokes will be ignored leading to failed authentication.
in order to handle such case, I will add an option to control the delay between keystrokes instead of hardcoding 100ms value. I will also try to get hold on a Yubikey to do tests on my side. I will update this thread with my progress.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Dear all,
I have been using VC with my new HP Spectre for a few months now and have been experiencing some strange behavior I would like to ask you about (system details see below).
I use a YubiKey with a fixed password plus a 4-digit pin. So the pre-boot sequence is (pin) + (yubikeypass) + (enter, from yubikey). No Pim.
It started to receive once in a while a “failed password, hash or pim” message on the first try only. Now I receive such a fail on first attempt on all boot processes. However, most of the time the second attempt works. Strangely the second attempt-fails are growing…. then third attempt works. So it seems I have “evolved” from 0 to 100% first attempt fails and now moving towards 100% 2nd attempt fails… (???)
Anyone an Idea what this is about and how I can fix it? So far I could always enter Windows, it is just annoying for the moment.
Thank you!
Michael
System:
HP Spectre x360 Convertible 13-ap0xxx, BIOS F.07
EFI Version 2.70, x64
VeraCrypt 1.23 (x64)
VeraCrypt Bootloader Config:
I've got a similar setup: HP Spectre 360 running Win 10 with Veracrypt v1.23 and using YubiKey to enter the pre-boot password.
My issue is that regardless of the password length (tried 64 and 32 char) I program into the key, VC only accepts the first 16-20 characters before moving on to the PIM (still not sure what that is) and the authorization fails. I can still enter the (very long, very complex) password manually for access.
The Yubikey dumps the complete password into a text editor just fine, so I'm prety sure its programmed correctly. Anyone know a reason why VC won't accept the entire password?
Perhaps @idrassi can help with this issue...
I'm having the same issue with a machine with EFI enabled. Despite using the YubiKey Personalization Tool to set a delay of 60ms, I'm experiencing the same issue as @msm1111 above. Once you're booted into Windows, everything is fine so it must be the EFI boot process that changed recently with VC 1.23.
I'm currently using VC 1.23 Hotfix 2 and from what I recall, I didn't have this issue pre VC 1.23. @idrassi, would it be possible to add a configurable key delay at the bootloader?
Thank you!
Its been couple of months. And i would like to refresh this topic, as it is still valid request. Can we get a confirmation that it will be soved one day or should we start looking for another soluton?
@msm1111, @jhaessly, @alancode, @fastriffs: sorry for the late answer. VeraCrypt EFI bootloader has a 100ms delay between accepted key strokes and so if a key received less than 100 ms before previous key then it will be ignored. This mechanism was added to fix issues with some keyboards that send keystores in double during a short period.
Unfortunately, it looks like yubikey has a maximum delay of 60ms between keystrokes, so most certainly some of the keystrokes will be ignored leading to failed authentication.
in order to handle such case, I will add an option to control the delay between keystrokes instead of hardcoding 100ms value. I will also try to get hold on a Yubikey to do tests on my side. I will update this thread with my progress.
How are we doing with this?
I am experiencing the same issue with a YubiKey 5 and VeraCrypt 1.26.7. using an Acer laptop.