Thread (15 messages) 15 messages, 6 authors, 2019-03-25

Re: [PATCH 0/2] Remove support for deprecated %pf and %pF in vsprintf

From: Sakari Ailus <sakari.ailus@linux.intel.com>
Date: 2019-03-22 13:54:01
Also in: ceph-devel, linux-acpi, linux-arm-kernel, linux-block, linux-btrfs, linux-f2fs-devel, linux-mm, linux-mmc, linux-pci, linux-pm, linux-scsi, linux-um, lkml, nvdimm, sparclinux

Hi Geert,

On Fri, Mar 22, 2019 at 02:37:18PM +0100, Geert Uytterhoeven wrote:
Hi Sakari,

On Fri, Mar 22, 2019 at 2:25 PM Sakari Ailus
[off-list ref] wrote:
quoted
The printk family of functions supports %ps and %pS conversion specifiers
to print function names. Yet the deprecated %pf and %pF conversion
specifiers with equivalent functionality remain supported. A number of
users of %pf and %pF remain.

This patchsets converts the existing users of %pf and %pF to %ps and %pS,
respectively, and removes support for the deprecated %pf and %pF.

The patches apply cleanly both on 5.1-rc1 as well as on Linux-next. No new
%pf or %pF users have been added in the meantime so the patch is
sufficient as itself on linux-next, too.
Do you know in which commit they became deprecated, so the backporters
know how far this can be backported safely?
That appears to be 04b8eb7a4ccd
("symbol lookup: introduce dereference_symbol_descriptor()"), the same
patch that made %p[fF] and %p[sS] functionally equivalent.

But my personal opinion would be not to backport the patch for two reasons:
the sheer number of files it touches (those format strings change for
various reasons) and the meager benefits it has on older kernels as any
backported patch using %s or %S still works as such. Porting a patch
forward should have no issues either as checkpatch.pl has been complaining
of the use of %pf and %pF for a while now.

-- 
Kind regards,

Sakari Ailus
sakari.ailus@linux.intel.com
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help