Thread (33 messages) 33 messages, 4 authors, 2025-07-05

Re: [PATCH v13 2/5] rust: support formatting of foreign types

From: Tamir Duberstein <hidden>
Date: 2025-07-03 22:46:30
Also in: dri-devel, linux-block, linux-clk, linux-devicetree, linux-kselftest, linux-pci, linux-pm, lkml, llvm, nouveau, rust-for-linux

On Thu, Jul 3, 2025 at 5:38 PM Miguel Ojeda
[off-list ref] wrote:
On Thu, Jul 3, 2025 at 11:28 PM Andrew Lunn [off-list ref] wrote:
quoted
A small patch tends to be more obviously correct than a big patch. The
commit message is more focused and helpful because it refers to a
small chunk of code. Because the commit message is more focused, it
can answer questions reviewers might ask, before they ask them. If i
Yeah, also better for smaller reverts, as well as typically easier to
backport if needed, etc.
I appreciate that this advice is well-intentioned, thank you. I agree
that all things being equal, small changes are better. In this
particular case, there are specific downsides to splitting for its own
sake which I tried to explain in previous replies: splitting the proc
macro from the rest of the machinery risks forcing a reviewer to
assess a chunk of code without seeing how it is used; in my experience
this limits the scope of the review. Splitting by subsystem has other
downsides, which I attempted to enumerate in my reply to Benno in the
other fork of this discussion (let's discuss those there, please).

There's also a tactical question about splitting by subsystem: are
there any tools that would assist in doing this, or is it a matter of
manually consulting MAINTAINERS to figure out file groupings?
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help