Re: [PATCH v10] show-branch: convert per-branch flags to commit-slab
Gatla Vishweshwar Reddy <[email protected]>
| Newsgroups | org.kernel.vger.git |
|---|---|
| Message-ID | <[email protected]> |
> We are removing this initialization. Why? Have we changed the > API to make BSS initialization sufficient? Does the updated > code no longer use this structure? Do we initialize it > somewhere else now? The init_commit_name_slab() call was removed by mistake in v7 when I restructured the initialization order. I moved the slab initialization for rev_flags_slab to after ref collection, and accidentally dropped the name_slab initialization entirely in the process. The API has not changed — BSS initialization is not sufficient because slab_size remains 0, causing division by zero when commit_to_name() tries to compute which slab page contains a commit's data using c->index / s->slab_size. The structure is still used throughout — name_commits(), name_commit(), name_parent() all call commit_to_name(). It is not initialized anywhere else. It was a plain mistake that I should have caught by running ./git show-branch master before sending. > Firing off a new iteration before there is a rough consensus > on what it should look like is a total waste of everyone's time. Understood. I will not send another version until I answer every question through discussion first. > If you can resurrect and continue the chat session with the > AI agent that spawned the v9 patch, ask it why it decided > to delete that init_commit_name_slab() call. The deletion happened when I restructured the initialization block in v7 — moving flags_stride and init_commit_rev_flags_with_stride to after ref collection. The name_slab init was in the same block and got dropped in the process. The AI did not flag it as significant because it was focused on the rev_flags_slab changes. I did not catch it because I did not run the binary. Both failures — the AI not flagging it and me not running it — are my failures to own. Please ask me any questions you need answered to be satisfied this is a human-understood patch. I will answer each and move this patch forward. Vishweshwar