b4 tag parsing too lax?

Hi,

I just saw this in commit ef6e1997da63 ("wifi: mac80211: fortify the spinlock against deadlock by interrupt"):

Reviewed-by tag, or it goes automatically?

I tracked it down to:

https://lore.kernel.org/all/[email protected]/

I can't tell if we should just deal with kind of unfortunate misparsing, or if b4 should require at least a "@" in the tag? (Though plenty of tags are just someone's name without an email in the cases where no email was available...)

Quoted 7 lines · Kees Cook

Hi,

I just saw this in commit ef6e1997da63 ("wifi: mac80211: fortify the spinlock against deadlock by interrupt"):

Reviewed-by tag, or it goes automatically?

Quoted 8 lines

I tracked it down to:

https://lore.kernel.org/all/[email protected]/

I can't tell if we should just deal with kind of unfortunate misparsing, or if b4 should require at least a "@" in the tag? (Though plenty of tags are just someone's name without an email in the cases where no email was available...)

I don't think b4 did this, because we already have precautions against this in place. E.g. if you retrieve that with b4, you will see that it doesn't get confused by that line:

$ b4 am -o/tmp https://lore.kernel.org/netdev/[email protected]/
Analyzing 5 messages in the thread
...
---
  ✗ [PATCH v4 1/1] wifi: mac80211: fortify the spinlock against deadlock by interrupt
    + Reviewed-by: Leon Romanovsky <[email protected]> (✓ DKIM/kernel.org)
  ---
  ...
---
...

While it's possible that b4 did this, it's not any current version.

Quoted 7 lines · Konstantin Ryabitsev

Quoted 6 lines · Kees Cook

Hi,

I just saw this in commit ef6e1997da63 ("wifi: mac80211: fortify the spinlock against deadlock by interrupt"):

Reviewed-by tag, or it goes automatically?

Quoted 1 line

While it's possible that b4 did this, it's not any current version.

Okay, weird! I have no idea how that got there then.

I should have CCed Johannes Berg to start with, but now I have. :) I'm curious what the workflow for that patch was -- do you use b4? (Perhaps this was a b4 behavior accidentally back in May?)

We use patchwork in wireless to assign the work between Kalle and I. I guess I could still apply patches using b4 though, but don't, at this point.

Quoted 8 lines · Kees Cook

Quoted 7 lines · Konstantin Ryabitsev

Quoted 6 lines · Kees Cook

Hi,

I just saw this in commit ef6e1997da63 ("wifi: mac80211: fortify the spinlock against deadlock by interrupt"):

Reviewed-by tag, or it goes automatically?

Heh, that's kinda funny.

Quoted 1 line

Quoted 1 line

While it's possible that b4 did this, it's not any current version.

It didn't.

Quoted 6 lines

Okay, weird! I have no idea how that got there then.

I should have CCed Johannes Berg to start with, but now I have. :) I'm curious what the workflow for that patch was -- do you use b4? (Perhaps this was a b4 behavior accidentally back in May?)

We use patchwork in wireless to assign the work between Kalle and I. I guess I could still apply patches using b4 though, but don't, at this point.

So I guess it's more a bug to report to patchwork, but I'm not sure how much good that will do unless we fix it ourselves :)