From: Kristoffer Haugsbakk <kristofferhaugsbakk@fastmail.com>
Date: Mon, 15 Dec 2025 12:04:54 GMT
Subject: Re: [PATCH] replay: move onto NULL check before first use
Message-ID: <a395825a-a9e9-4cde-bf2d-f9b72de9212d@app.fastmail.com>
In-Reply-To: <a017e50f-7c8f-461f-8627-2fd1445d29f6@gmail.com>

On Mon, Dec 15, 2025, at 11:10, Phillip Wood wrote:
> On 11/12/2025 17:56, René Scharfe wrote:
>> cmd_replay() aborts if the pointer "onto" is NULL after argument
>> parsing, e.g. when specifying a non-existing commit with --onto.
>> 15cd4ef1f4 (replay: make atomic ref updates the default behavior,
>> 2025-11-06) added code that dereferences this pointer before the check.
>> Switch their places to avoid a segmentation fault.
>
> This fixes the regression nicely. There is a preexisting bug that we
> treat an invalid --onto argument the same as a missing argument but that
> can be fixed separately.

I have a commit cooking (locally) which makes the command die when it
cannot find commit-ish for `--onto` or `--advance` (whitespace mangled
diff):

```
diff --git a/builtin/replay.c b/builtin/replay.c
index 507b909df7d..72d62aa34a6 100644
--- a/builtin/replay.c
+++ b/builtin/replay.c
@@ -39,7 +39,7 @@ static struct commit *peel_committish(struct repository *repo, const char *name)
 	struct object_id oid;

 	if (repo_get_oid(repo, name, &oid))
-		return NULL;
+		die(_("'%s' is not a valid commit-ish"), name);
 	obj = parse_object(repo, &oid);
 	return (struct commit *)repo_peel_to_type(repo, name, 0, obj,
 						  OBJ_COMMIT);
```

Instead of dieing like this:

    Replaying down to root commit is not supported yet!

I hope that doesn’t cause any leak issues when I test it later.