packetdrill: add test of tcp window clamp socket option

[nspring]

Adds a test of TCP_WINDOW_CLAMP, including lowering, raising, and
lowering again the clamp while the connection is in progress.

Requires a patch to tcp_set_window_clamp, since prior linux ignores
this field unless set before the connection exchanges data.

Signed-off-by: Neil Spring [email protected]

[google-cla]

Thanks for your pull request. It looks like this may be your first contribution to a Google open source project (if not, look below for help). Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

📝 Please visit https://cla.developers.google.com/ to sign.

Once you've signed (or fixed any issues), please reply here with @googlebot I signed it! and we'll verify it.

---

What to do if you already signed the CLA

Individual signers

• It's possible we don't have your GitHub username or you're using a different email address on your commit. Check your existing CLA data and verify that your email is set on your git commits.

Corporate signers

• Your company has a Point of Contact who decides which employees are authorized to participate. Ask your POC to be added to the group of authorized contributors. If you don't know who your Point of Contact is, direct the Google project maintainer to go/cla#troubleshoot (Public version).
• The email used to register you as an authorized contributor must be the email used for the Git commit. Check your existing CLA data and verify that your email is set on your git commits.
• The email used to register you as an authorized contributor must also be attached to your GitHub account.

ℹ️ Googlers: Go here for more info.

[nealcardwell]

Hi, Neil! Good to hear from you. Still tuning receive windows after all these years! I love it! ;-)

And a very nice test!

For extra clarity, do you mind putting this in a directory called "window_clamp" rather than "clamp"?

And to help coordinate between your kernel patch and this test, do you mind putting in the commit message the title of the patch you have posted or are planning to post that will make this test pass?

Thanks!

[nealcardwell]

Thanks! Do you mind putting this after TCP_DEFER_ACCEPT and before TCP_INFO? That's where we have it in our internal sources. (And somehow that line didn't make it upstream... sorry!)

[nealcardwell]

When I look at:
https://github.com/google/packetdrill/pull/56/commits
it says:
nspring wants to merge 3 commits into google:master from nspring:window_clamp

Can you please squash/fixup the 3 commits into a single commit? Thanks!

[nspring]

Gotcha - commits squashed, file/directory renamed, and a bit more test code added to exercise a scenario Eric Dumazet pointed out in the discussion on the kernel patch. I'm not quite sure the patch is perfect yet; maybe it's too conservative when raising the clamp.

I'm really happy you thought the test was good. I wonder if you can help me with two things though.

1. BSD? if run without linux, will the test explode because window clamp is not defined?
2. Am I being too specific about the advertised window values on every ack? I don't care about precise intermediate advertised windows, only that it eventually reaches the right value and maybe that it doesn't increase or decrease when it shouldn't. I lean toward including them to document what's happening, but don't want an off-by-one rounding error to make someone else have to update the test.

Still tuning receive windows after all these years!

Low hanging fruit! Nice to be able to work together (even if indirectly) again; this is a sweet project.