mbox series

[0/9] remove some of unneeded kallsyms includes

Message ID 20171208025616.16267-1-sergey.senozhatsky@gmail.com
Headers show
Series remove some of unneeded kallsyms includes | expand

Message

Sergey Senozhatsky Dec. 8, 2017, 2:56 a.m. UTC 
Hello,

	A small patch set that removes some kallsyms includes
here and there. Mostly those kallsyms includes are leftovers:
printk() gained %pS/%pF modifiers support some time ago, so
print_symbol() and friends became sort of unneeded [along with
print_fn_descriptor_symbol() deprecation], thus some of the
users were converted to pS/pF. This patch set just cleans up
that convertion.

	We still have a number of print_symbol() users [which
must be converted to ps/pf, print_symbol() uses a stack buffer
KSYM_SYMBOL_LEN to do what printk(ps/pf) can do], but this is
out of scope.

	I compile tested the patch set; but, as always and
usual, would be great if 0day build robot double check it.

Sergey Senozhatsky (9):
  sched/autogroup: remove unneeded kallsyms include
  mm: remove unneeded kallsyms include
  power: remove unneeded kallsyms include
  pci: remove unneeded kallsyms include
  pnp: remove unneeded kallsyms include
  mm: remove unneeded kallsyms include
  workqueue: remove unneeded kallsyms include
  hrtimer: remove unneeded kallsyms include
  genirq: remove unneeded kallsyms include

 drivers/base/power/main.c | 1 -
 drivers/pci/quirks.c      | 1 -
 drivers/pnp/quirks.c      | 1 -
 kernel/irq/spurious.c     | 1 -
 kernel/sched/autogroup.c  | 1 -
 kernel/time/hrtimer.c     | 1 -
 kernel/workqueue.c        | 1 -
 mm/memory.c               | 4 ----
 mm/vmalloc.c              | 1 -
 9 files changed, 12 deletions(-)

Comments

Andrew Morton Dec. 8, 2017, 10:24 p.m. UTC | #1 
On Fri,  8 Dec 2017 11:56:07 +0900 Sergey Senozhatsky <sergey.senozhatsky.work@gmail.com> wrote:

> 	A small patch set that removes some kallsyms includes
> here and there. Mostly those kallsyms includes are leftovers:
> printk() gained %pS/%pF modifiers support some time ago, so
> print_symbol() and friends became sort of unneeded [along with
> print_fn_descriptor_symbol() deprecation], thus some of the
> users were converted to pS/pF. This patch set just cleans up
> that convertion.
> 
> 	We still have a number of print_symbol() users [which
> must be converted to ps/pf, print_symbol() uses a stack buffer
> KSYM_SYMBOL_LEN to do what printk(ps/pf) can do], but this is
> out of scope.
> 
> 	I compile tested the patch set; but, as always and
> usual, would be great if 0day build robot double check it.

I grabbed everything and shall drop any patches which later turn up in
the various subsystem trees.
Sergey Senozhatsky Dec. 9, 2017, midnight UTC | #2 
On (12/08/17 14:24), Andrew Morton wrote:
> On Fri,  8 Dec 2017 11:56:07 +0900 Sergey Senozhatsky <sergey.senozhatsky.work@gmail.com> wrote:
> 
> > 	A small patch set that removes some kallsyms includes
> > here and there. Mostly those kallsyms includes are leftovers:
> > printk() gained %pS/%pF modifiers support some time ago, so
> > print_symbol() and friends became sort of unneeded [along with
> > print_fn_descriptor_symbol() deprecation], thus some of the
> > users were converted to pS/pF. This patch set just cleans up
> > that convertion.
> > 
> > 	We still have a number of print_symbol() users [which
> > must be converted to ps/pf, print_symbol() uses a stack buffer
> > KSYM_SYMBOL_LEN to do what printk(ps/pf) can do], but this is
> > out of scope.
> > 
> > 	I compile tested the patch set; but, as always and
> > usual, would be great if 0day build robot double check it.
> 
> I grabbed everything and shall drop any patches which later turn up in
> the various subsystem trees.

thank you!

	-ss