Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

sha256crypt-ztex #3374

Merged
merged 1 commit into from Aug 18, 2018
Merged

sha256crypt-ztex #3374

merged 1 commit into from Aug 18, 2018

Conversation

Apingis
Copy link
Collaborator

@Apingis Apingis commented Aug 18, 2018

sha256crypt-ztex

  • incl. Verilog sources
  • Design tools reported possible frequency is 166.3. At that frequency tested boards fail
    often (timeout) and lose successful guesses. Default frequency is set to 135.
  • Performance: 71.5 Kc/s for length 8; 60.6 Kc/s for length 12.
  • Measured power consumption (12V): ~2.6 А. Idle power consumption is ~1.0 А.

@kholia kholia requested a review from solardiz Aug 18, 2018
Copy link
Member

@solardiz solardiz left a comment

I've made some comments on improving the documentation, which should be done with separate commits/PR. I'll merge this now, then test this PR's new functionality after merging.

| | / \ /
| process_bytes | / "main" \ <------
| |<--+--| memory |
+---------------+ | \ (16x256b) / <------
Copy link
Member

@solardiz solardiz Aug 18, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Should be 6x128B. BTW, let's use "B" for bytes if we use "b" for bits (ditto for the sha512crypt-ztex README). These documentation issues can be fixed in a separate commit and PR for both sha256crypt-ztex and sha512crypt-ztex.

| procb_buf | -->| C.P.U. | |
+-----------+ | +-----------------------+ | |
^ | | integer ops. incl. | | |
\ | | 6 x 16 registers(x16b)| | |
Copy link
Member

@solardiz solardiz Aug 18, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Here's where we use "b" for bits, and that's OK.


- Each FPGA has 25 computing units, that's 75 cores, 150 keys are
computed in parallel.

Copy link
Member

@solardiz solardiz Aug 18, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It would be helpful to add (with the separate commit/PR) some info on relative resource usage by SHA-256 cores vs. the "overhead" (CPUs, glue logic). We do have this info in sha512crypt-ztex's README.

@@ -376,6 +376,14 @@ Frequency = 141
# This turns off all 10 units (resulting in a timeout).
#Config1_04A36E0FD6_0 = \xff\x03

[ZTEX:sha256crypt]
# Design tools reported possible frequency is 166.3 but tested boards
# miss guesses, often fail unless frequency is decreased.
Copy link
Member

@solardiz solardiz Aug 18, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It'd be helpful to add a similar comment for sha512crypt and Drupal7, stating those formats' design tools' reported frequency.

Frequency = 135
#TargetRounds = 1000000


Copy link
Member

@solardiz solardiz Aug 18, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Extra linefeed here.

Copy link
Collaborator Author

@Apingis Apingis Aug 18, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you!
The documentation definitely requires an update / correction. I'm preparing a PR to include topics you've commented on. Maybe some other items require more detailed description, such as resource usage, performance of 16-bit CPU in sha256crypt-ztex vs 32-bit CPU in sha512crypt-ztex.
Regarding extra linefeed in run/john.conf:L386, I was unsure on style in john.conf. Sometimes sections that refer to different subsystems are separated with 2 empty lines, i.e. L419, L474.

Copy link
Member

@solardiz solardiz Aug 18, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

You're right about the extra linefeed, I missed the fact this was the very last ZTEX entry, and is followed by OpenCL stuff. Makes sense.

@solardiz solardiz merged commit d2b95cd into openwall:bleeding-jumbo Aug 18, 2018
2 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

2 participants