Add control command to drop guard nodes

The attached patches add a new control command to Tor "DUMPGUARDS" that tells the server to remove all guard nodes.

changed milestone to %Tor: 0.2.5.x-final

added component::core tor/tor controller guards milestone::Tor: 0.2.5.x-final priority::low resolution::implemented status::closed tor-client type::enhancement labels

Trac:
0001-add-DUMPGUARDS-command.patch

patch to control-spec.txt

Trac:
tor-dumpguards.patch

Patch to Tor source

Quick notes:

• Should there be an ability to drop a specific guard?
• Right now, if you call this command with arguments, they all get ignored. That can't be right.
• The loop is strangely structured, given that "i" is never incremented. Perhaps it would make more sense to do while (smartlist_len(entryguards) { entryguard = smartlist_get(entryguards, 0); ... }?
• Remember that in C, "void fn()" is a function that takes an unspecified number of arguments. If you need to write a function that takes no arguments, you have to say "void fn(void)".
• Every function should be documented.
• Perhaps we should explain briefly what the intended use case of this command is.

Trac:
Keywords: N/A deleted, tor-client guards controller added
Status: new to needs_revision
Milestone: N/A to Tor: 0.2.5.x-final

Replying to nickm:

• Should there be an ability to drop a specific guard?

Dropping all guards is sufficient for my use case and the one expressed on tor-dev but I can't give a general answer on that.

Replying to nickm:

• Right now, if you call this command with arguments, they all get ignored. That can't be right.
• The loop is strangely structured, given that "i" is never incremented. Perhaps it would make more sense to do while (smartlist_len(entryguards) { entryguard = smartlist_get(entryguards, 0); ... }?
• Remember that in C, "void fn()" is a function that takes an unspecified number of arguments. If you need to write a function that takes no arguments, you have to say "void fn(void)".

I will rewrite and resubmit the patch.

• Every function should be documented.

I though that documenting "remove_all_entry_guards" is superfluous but I will add a short documentation to it.

• Perhaps we should explain briefly what the intended use case of this command is.

To the control-spec.txt..?

Trac:
tor-dumpguards.2.patch

v2

Almost: that one is doing a string length comparison, so it will reject the command if there's space after it which probably isn't what we wanted. (Other commands allow additional space.)

I'll see if it's easy to just tweak that up and merge it.

Trac:
Status: needs_revision to needs_review

I had to tweak it a little to make it merge into master. How do you like branch "bug9934_nm" in my public repository ( https://gitweb.torproject.org/nickm/tor.git ) ?

I also renamed the command to DROPGUARDS, since "dump" in programmerese can mean either "discard" or "print".

Please let me know if you like this version of the patch, and whether I should credit the patch to "ra" or to some longer name.

Replying to nickm:

I had to tweak it a little to make it merge into master. How do you like branch "bug9934_nm" in my public repository ( https://gitweb.torproject.org/nickm/tor.git ) ?

The smartlist solution seems cleaner than the string length comparison. I tried but couldn't figure how to do it in a short time. Anyways, the branch looks perfectly fine to me.

I also renamed the command to DROPGUARDS, since "dump" in programmerese can mean either "discard" or "print".

Ack.

Please let me know if you like this version of the patch, and whether I should credit the patch to "ra" or to some longer name.

"ra" is fine.

Okay, merged. Thanks!

Trac:
Status: needs_review to closed
Resolution: N/A to implemented

closed

moved to tpo/core/tor#9934 (closed)