[Bug 221696] btmtk: regression in 6.6.142: NULL pointer dereference in btmtk_usb_hci_wmt_sync during resume from S4

[bugzilla-daemon@kernel.org]

https://bugzilla.kernel.org/show_bug.cgi?id=221696

--- Comment #8 from Matt Whitlock (kernel@mattwhitlock.name) ---
I've figured out where the bisection went wrong. It isn't sufficient to tell
Git that a commit that exhibits the bug is "old" and one that doesn't is "new"
because git-bisect actually wants to test some commits that do not have the
bug-introducing commit as an ancestor.

So, at each step of the bisection, I have to manually check that the
bug-introducing commit (d019930b0049fc2648a6b279893d8ad330596e81) is actually
an ancestor of the current bisection step, and if it isn't, then the current
step is definitionally "old":

$ git merge-base --is-ancestor \
      d019930b0049fc2648a6b279893d8ad330596e81 BISECT_HEAD ||
  git bisect old

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are the assignee for the bug.