On Sat, Jan 19, 2013 at 3:01 PM, Rafael J. Wysocki [off-list ref] wrote:

On Saturday, January 19, 2013 02:50:17 PM Sedat Dilek wrote:

quoted

On Sat, Jan 19, 2013 at 2:45 PM, Rafael J. Wysocki [off-list ref] wrote:

quoted

On Friday, January 18, 2013 05:42:20 PM Randy Dunlap wrote:

quoted

On 01/17/13 20:37, Stephen Rothwell wrote:

quoted

Hi all,

Changes since 20130117:

on x86_64:

  CC      drivers/acpi/device_pm.o
drivers/acpi/device_pm.c:778:5: error: redefinition of 'acpi_dev_suspend_late'
In file included from include/acpi/acpi_drivers.h:29:0,
                 from drivers/acpi/device_pm.c:33:
include/linux/acpi.h:526:19: note: previous definition of 'acpi_dev_suspend_late' was here
drivers/acpi/device_pm.c:810:5: error: redefinition of 'acpi_dev_resume_early'
In file included from include/acpi/acpi_drivers.h:29:0,
                 from drivers/acpi/device_pm.c:33:
include/linux/acpi.h:527:19: note: previous definition of 'acpi_dev_resume_early' was here
drivers/acpi/device_pm.c:828:5: error: redefinition of 'acpi_subsys_prepare'
In file included from include/acpi/acpi_drivers.h:29:0,
                 from drivers/acpi/device_pm.c:33:
include/linux/acpi.h:528:19: note: previous definition of 'acpi_subsys_prepare' was here
drivers/acpi/device_pm.c:846:5: error: redefinition of 'acpi_subsys_suspend_late'
In file included from include/acpi/acpi_drivers.h:29:0,
                 from drivers/acpi/device_pm.c:33:
include/linux/acpi.h:529:19: note: previous definition of 'acpi_subsys_suspend_late' was here
drivers/acpi/device_pm.c:861:5: error: redefinition of 'acpi_subsys_resume_early'
In file included from include/acpi/acpi_drivers.h:29:0,
                 from drivers/acpi/device_pm.c:33:
include/linux/acpi.h:530:19: note: previous definition of 'acpi_subsys_resume_early' was here
make[3]: *** [drivers/acpi/device_pm.o] Error 1




Full randconfig file is attached.

Thanks, I'll take it for inclusion into the build testing framework.

The appended patch fixes the problem for me.

Thanks,
Rafael


---
From: Rafael J. Wysocki <redacted>
Subject: ACPI / PM: Fix build for unusual combination of Kconfig options

CONFIG_PM_SLEEP may be set even if CONFIG_ACPI_SLEEP is unset,
although that is unusual.  For this reason, make the headers of
functions built for both CONFIG_ACPI and CONFIG_PM_SLEEP set
simultaneously depend on that combination of Kconfig options
instead of CONFIG_ACPI_SLEEP.

This fixes a build problem reported by Randy Dunlap.

Signed-off-by: Rafael J. Wysocki <redacted>
---
 include/linux/acpi.h |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Index: linux-pm/include/linux/acpi.h
===================================================================

--- linux-pm.orig/include/linux/acpi.h
+++ linux-pm/include/linux/acpi.h

Can you tell me how do you create such a patch with "Index:" line?

linux-pm.orig as a reference is not very meaninful :-).

If this is against Linux-Next (next-20130118) I would like to see:

--- next-20130118.orig/path/to/file
+++ next-20130118/path/to/file

I had used 'git format-patch' with '--subject-prefix="PATCH
next-20130118"' option in such a case.
Just as a hint by not telling you how you should do your job.

/me is still a Git n00b!

Well, I use quilt to generate patches. :-)

This particular one was generated with "quilt refresh --diffstat", I'm not
sure how to get the same result using git, sorry.

OK, I know I am a bad Git teacher, but from my p00r Git experiences...

I required some of the below options of 'git format-patch' for the
Freetz router project.

[ Fight with "p0" VS. "p1" patch-format aka w/ or w/o prefix ]

In general Freetz wants its patches in "p0" patch-format to fit its
own "freetz_patch" tool.
( When I asked on the Git ML people did not know/like the terms "p0"
and "p1" so it's in parenthesis. )
This means patches should NOT contain "a/" and "b/" prefixes.

       --no-prefix
           Do not show any source or destination prefix.

[ Add a "source" (override "a/") and a "destination" (override "b/") prefix ]

You will like this one :-).
In combination with the above I had to add
"--src-prefix=linux-2.6.13.1.orig" and -dst-prefix="linux-2.6.13.1"
for kernel-patches.
YES, my router has a very ancient kernel-release :-(!

       --src-prefix=<prefix>
           Show the given source prefix instead of "a/".

       --dst-prefix=<prefix>
           Show the given destination prefix instead of "b/".

[ Do NOT number patches in a series ]

And last but not least, I personally do not like to see "[PATCH]
12/23" in a series of patches.
Suppress this by using "--no-numbered" option.
NOTE: For a single patch this option is useless :-)!

       -N, --no-numbered
           Name output in [PATCH] format.

[ bonbon: cover-letter aka 0/x or 00/yy or 000/zzz (IMHO zzz is more a
"love-letter" :-)) ]

       --cover-letter
           In addition to the patches, generate a cover letter file
containing the shortlog and the overall diffstat. You can fill in a
description in the file before sending it out.

I like when people send a cover-letter where they put all the relevant
informations for the ***complete*** patch-series.
This is very helpful for people who want to review.
So, you can reply comments to the cover-letter-email for a general discussions.
Comments on the single patches you can do appropriately.

A book recommendation for German speaking people (see [1])...
( IMPORTANT NOTE: I get no money for doing marketing for the book! )

Hope this helps you :-)!
But make your own Git experiences!

- Sedat -

[1] https://www.opensourcepress.de/git/

Thanks,
Rafael


--
I speak only for myself.
Rafael J. Wysocki, Intel Open Source Technology Center.