[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20211011080135.GA11242@lst.de>
Date: Mon, 11 Oct 2021 10:01:35 +0200
From: Christoph Hellwig <hch@....de>
To: Simon Ser <contact@...rsion.fr>
Cc: Christoph Hellwig <hch@....de>,
Alex Deucher <alexdeucher@...il.com>,
Stephen Rothwell <sfr@...b.auug.org.au>,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
Linux Next Mailing List <linux-next@...r.kernel.org>,
Daniel Vetter <daniel@...ll.ch>,
Benson Leung <bleung@...omium.org>,
Enric Balletbo i Serra <enric.balletbo@...labora.com>,
Greg Kroah-Hartman <gregkh@...uxfoundation.org>
Subject: Re: linux-next: build failure after merge of the amdgpu tree
On Mon, Oct 11, 2021 at 07:57:17AM +0000, Simon Ser wrote:
> No, we can't have a "I_AM_NOT_BROKEN" ioctl for each and every uAPI mis-use.
> User-space detection has been determined to be the best course of action.
If your API addition breaks userspace, yes you need an add-in.
> I guess I'll just inline these functions in the driver then, if a revert will
> be NACK'ed by you?
It will be NACKed, and I will also complain to Linus about any PR containing
buggy code like this.
With your completely broken change you cement in a mapping of an executable
name to map to what you consider a "bug" without any way to fix it up.
Which is even worse for a something fast moving like chrome/chromeos which
will eventually gets its act together and fix things while you'll keep a
weird feature mismatch just for it around forever.
No wonder our graphics stack is stuch a convoluted buggy mess if you keep
piling broken workarounds over workarounds instead of sorting things out
properly.
Powered by blists - more mailing lists