This reverts commit 953742c8.

Dumping two lines into autoconf.h for all existing Kconfig options
results in a giant file (~16k lines) we have to process each time we
compile something.  We've weaned IS_ENABLED() and similar off of
requiring the __enabled_ definitions so now we can revert the change
which caused all the extra lines.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Provide a -r option to display when fragments contain redundant
options. This is really useful when breaking apart a config into
fragments, as well as cleaning up older fragments.
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>
Acked-by: NDarren Hart <dvhart@linux.intel.com>
Acked-by: NBruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

Somehow the merge_config.sh script didn't get its execute bit
set when it was merged. Fix this.
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>
Acked-by: NDarren Hart <dvhart@linux.intel.com>
Acked-by: NBruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

Take the first config fragment and use it verbatim as the initial config
set.  This avoids running the verification loop for the first file, as
nothing has actually been merged at this point.  This significantly
increases performance for large config fragments.
Signed-off-by: NJosh Boyer <jwboyer@redhat.com>
Acked-by: NJohn Stultz <john.stultz@linaro.org>
Acked-by: NDarren Hart <dvhart@linux.intel.com>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

Running `oldconfig' after any of the following configuration change:

either trivial addition, such as:

config A
	bool "A"

choice
	prompt "Choice ?"
	depends on A

	config CHOICE_B
		bool "Choice B"

	config CHOICE_C
		bool "Choice C"
endchoice

or more tricky change:

OLD KCONFIG                      |  NEW KCONFIG
                                 |
                                 |  config A
                                 |          bool "A"
                                 |
choice                           |  choice
        prompt "Choice ?"        |          prompt "Choice ?"
                                 |
        config CHOICE_C          |          config CHOICE_C
                bool "Choice C"  |                  bool "Choice C"
                                 |
        config CHOICE_D          |          config CHOICE_D
                bool "Choice D"  |                  bool "Choice D"
endchoice                        |
                                 |          config CHOICE_E
                                 |                  bool "Choice E"
                                 |                  depends on A
                                 |  endchoice

will not cause the choice to be considered as NEW, and thus not be
asked. The cause of this behavior is that choice's novelty are computed
statically right after the saved configuration has been read. At this
point, the new dependency's value is still unknown and asserted to be
`no'. Moreover, no update to this decision is made afterward.

Correct this by dynamically evaluating a choice's novelty, and removing the
static evaluation.
Reported-and-tested-by: NUwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: NArnaud Lacombe <lacombar@gmail.com>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

This commit fixes a bug, while introducing a new one..

commit 7203ddbd4be9720649e47d756a001e0c7d7f8ae2
Author: Wang YanQing <udknight@gmail.com>
Date:   Thu Jan 12 11:31:32 2012 +0800

    menuconfig: let make not report error when not save configuration

Pressing ESC should cancel the yes/no dialog and return back to
the main menu, but not exit from menuconfig.
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Acked-by: NWang YanQing <udknight@gmail.com>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

Reinhard Tartler discovered a corner case of calling xfwrite() where the
length of the string is zero.

Arnaud Lacombe suggested to use assertion for the corner case, as
fwrite(3) is currently used:

 1) in comment printers. Empty comment are not allowed.
 2) in a callback passed to expr_print(), where the string printed is
    either NULL OR non-empty.
 3) in the lexer, auto-generated, and unused.

I feel using assertion is a good solution:

 1) It cleanly takes care of the above-mentioned corner case.
 2) It can be easily disabled by defining NDEBUG.
 3) It asserts xfwrite() is simply a wrapper for fwrite().
Reported-by: NReinhard Tartler <Reinhard.Tartler@informatik.uni-erlangen.de>
Signed-off-by: NArnaud Lacombe <lacombar@gmail.com>
Signed-off-by: NJean Sacren <sakiwit@gmail.com>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

Make the V=0 output from update-po-config be aligned correctly.
Also remove an outdated comment and add a "GEN" statement.
Signed-off-by: NPeter Foley <pefoley2@verizon.net>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

I find every time when I choice the 'NO' button at the dialog
which let me choice whether to save the configuration before exit
menuconfig, it always report the blow:

" GEN     /mnt/sda7/home/build/test/Makefile
  HOSTCC  scripts/kconfig/mconf.o
  HOSTLD  scripts/kconfig/mconf
scripts/kconfig/mconf Kconfig

Your configuration changes were NOT saved.

make[2]: *** [menuconfig] Error 1
make[1]: *** [menuconfig] Error 2
make: *** [sub-make] Error 2 "

This patch repair it.
Signed-off-by: NWang YanQing <udknight@gmail.com>
Acked-by: NDavidlohr Bueso <dave@gnu.org>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

Arnaud Lacombe pointed out the final checking that the requested configs
were included in the final .config was broken.

The example was that if you had a fragment that disabled
CONFIG_DECOMPRESS_GZIP applied to a normal defconfig, there would be no
final warning that CONFIG_DECOMPRESS_GZIP was acutally set in the final
.config.

This bug was introduced by me in v3 of the original patch, and the
following patch reverts the invalid change.
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>
Reported-by: NArnaud Lacombe <lacombar@gmail.com>
Cc: Darren Hart <dvhart@linux.intel.com>
Cc: Michal Marek <mmarek@suse.cz>
Cc: Arnaud Lacombe <lacombar@gmail.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

Fix whitespace usage in the clean_up routine.
Signed-off-by: NDarren Hart <dvhart@linux.intel.com>
Acked-by: NJohn Stultz <john.stultz@linaro.org>
Cc: Michal Marek <mmarek@suse.cz>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

The SIGHUP SIGINT and SIGTERM names caused failures when running
merge_config.sh with the dash shell.  Dropping the "SIG" component makes
the script work in both bash and dash.
Signed-off-by: NDarren Hart <dvhart@linux.intel.com>
Acked-by: NJohn Stultz <john.stultz@linaro.org>
Cc: Michal Marek <mmarek@suse.cz>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

After noticing almost every distro has their own method of managing config
fragments, I went looking at some best practices, and wanted to try to
consolidate some of the different approaches so this fairly simple
infrastructure can be shared (and new distros/build systems don't have to
implement yet another config fragment merge script).

This script is most influenced by the Windriver tools used in the Yocto
Project, reusing some portions found there.

This script merges multiple config fragments, warning on any overridden
values.  It then sets any unspecified values to their default, then
finally checks to make sure no specified value was dropped due to
unsatisfied dependencies.

I'm sure this implementation won't work for everyone, and I expect it will
need to evolve to adapt for various use cases.  But I think its a
reasonable starting point.

Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: Greg Thelen <gthelen@google.com>
Cc: Reinhard Tartler <Reinhard.Tartler@informatik.uni-erlangen.de>
Cc: Dmitry Fink <Dmitry.Fink@palm.com>
Cc: Darren Hart <dvhart@linux.intel.com>
Cc: Eric B Munson <ebmunson@us.ibm.com>
Cc: Bruce Ashfield <Bruce.Ashfield@windriver.com>
Cc: Michal Marek <mmarek@suse.cz>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

Thomas Lange reported that when he did a 'make localmodconfig', his
config was missing the brcmsmac driver, even though he had the module
loaded.

Looking into this, I found the file:
drivers/net/wireless/brcm80211/brcmsmac/Makefile
had the following in the Makefile:

MODULEPFX := brcmsmac

obj-$(CONFIG_BRCMSMAC)  += $(MODULEPFX).o

The way streamline-config.pl works, is parsing all the
 obj-$(CONFIG_FOO) += foo.o
lines to find that CONFIG_FOO belongs to the module foo.ko.

But in this case, the brcmsmac.o was not used, but a variable in its place.

By changing streamline-config.pl to remember defined variables in Makefiles
and substituting them when they are used in the obj-X lines, allows
Thomas (and others) to have their brcmsmac module stay configured
when it is loaded and running "make localmodconfig".
Reported-by: NThomas Lange <thomas-lange2@gmx.de>
Tested-by: NThomas Lange <thomas-lange2@gmx.de>
Cc: Arend van Spriel <arend@broadcom.com>
Cc: stable@vger.kernel.org
Signed-off-by: NSteven Rostedt <rostedt@goodmis.org>

Simplify the way lines ending with backslashes (continuation) in Makefiles
is parsed. This is needed to implement a necessary fix.
Tested-by: NThomas Lange <thomas-lange2@gmx.de>
Cc: stable@vger.kernel.org
Signed-off-by: NSteven Rostedt <rostedt@goodmis.org>

Commit 5c48b108 ("um: take arch/um/sys-x86 to arch/x86/um") broke the
make target update-po-config, as its symlink trick (again) fails.
(Previous breakage was fixed with commit bdc69ca4 ("kconfig: change
update-po-config to reflect new layout of arch/um").)

The new UML layout allows to drop the symlick trick entirely. And if,
one day, another architecture supports UML too, that should now work
without again breaking this make target.
Signed-off-by: NPaul Bolle <pebolle@tiscali.nl>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

Use the xfwrite wrapper function defined in lkc.h to check the return value of
fwrite and silence these warnings.

  HOSTCC  scripts/kconfig/zconf.tab.o
scripts/kconfig/zconf.tab.c: In function 'header_print_comment':
/usr/src/lto/scripts/kconfig/confdata.c:551:10: warning: ignoring return value of 'fwrite', declared with attribute warn_unused_result
scripts/kconfig/zconf.tab.c: In function 'kconfig_print_comment':
/usr/src/lto/scripts/kconfig/confdata.c:467:10: warning: ignoring return value of 'fwrite', declared with attribute warn_unused_result
Signed-off-by: NPeter Foley <pefoley2@verizon.net>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

Remove set but not used variables to fix warnings.

  HOSTCC  scripts/kconfig/gconf.o
/usr/src/lto/scripts/kconfig/gconf.c: In function 'change_sym_value':
/usr/src/lto/scripts/kconfig/gconf.c:833:11: warning: variable 'oldval' set but not used [-Wunused-but-set-variable]
/usr/src/lto/scripts/kconfig/gconf.c: In function 'update_tree':
/usr/src/lto/scripts/kconfig/gconf.c:1281:19: warning: variable 'prop' set but not used [-Wunused-but-set-variable]
Signed-off-by: NPeter Foley <pefoley2@verizon.net>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

Specify format arguments to fix warnings.

  HOSTCC  scripts/kconfig/gconf.o
/usr/src/lto/scripts/kconfig/gconf.c: In function 'on_introduction1_activate':
/usr/src/lto/scripts/kconfig/gconf.c:686:6: warning: format not a string literal and no format arguments
/usr/src/lto/scripts/kconfig/gconf.c: In function 'on_about1_activate':
/usr/src/lto/scripts/kconfig/gconf.c:704:6: warning: format not a string literal and no format arguments
/usr/src/lto/scripts/kconfig/gconf.c: In function 'on_license1_activate':
/usr/src/lto/scripts/kconfig/gconf.c:723:6: warning: format not a string literal and no format arguments
Signed-off-by: NPeter Foley <pefoley2@verizon.net>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

to make it easier to locate begin/end when editing long strings;
Signed-off-by: NCheng Renquan <crquan@gmail.com>
Acked By: Nir Tzachar <nir.tzachar@gmail.com>

The original dialog_inputbox doesn't work with longer than prompt_width
strings, here fixed it in this way:
1) add variable cursor_form_win to record cursor of form_win,
   keep its value always between [0, prompt_width-1];
   reuse the original cursor_position as cursor of the string result,
   use (cursor_position-cursor_form_win) as begin offset to show part of
   the string in form_win;
Signed-off-by: NCheng Renquan <crquan@gmail.com>
Cc: Arnaud Lacombe <lacombar@gmail.com>
Cc: Nir Tzachar <nir.tzachar@gmail.com>

To support unlimited length string config items;

No check for realloc return value keeps code simple, and to be
consistent with other existing unchecked malloc in kconfig.
Signed-off-by: NCheng Renquan <crquan@gmail.com>
Signed-off-by: NArnaud Lacombe <lacombar@gmail.com>

In case KEY_BACKSPACE / KEY_DC to delete a char, it memmove only
(len-cursor_position+1) bytes;
the default case is to insert a char, it should also memmove exactly
(len-cursor_position+1) bytes;

the original use of (len+1) is wrong and may access following memory
that doesn't belong to result, may cause SegFault in theory;

	case KEY_BACKSPACE:
		if (cursor_position > 0) {
			memmove(&result[cursor_position-1],
					&result[cursor_position],
					len-cursor_position+1);
			cursor_position--;
		}
		break;
	case KEY_DC:
		if (cursor_position >= 0 && cursor_position < len) {
			memmove(&result[cursor_position],
					&result[cursor_position+1],
					len-cursor_position+1);
		}
		break;
	default:
		if ((isgraph(res) || isspace(res)) &&
				len-2 < result_len) {
			/* insert the char at the proper position */
			memmove(&result[cursor_position+1],
					&result[cursor_position],
					len-cursor_position+1);
			result[cursor_position] = res;
			cursor_position++;
		}
Signed-off-by: NCheng Renquan <crquan@gmail.com>
Acked-by: NNir Tzachar <nir.tzachar@gmail.com>

Signed-off-by: NCheng Renquan <crquan@gmail.com>
Acked-by: NArnaud Lacombe <lacombar@gmail.com>

Some variables were being set but never used, which was triggering
warnings in GCC >= 4.6.
Signed-off-by: NLucas De Marchi <lucas.demarchi@profusion.mobi>
Signed-off-by: NArnaud Lacombe <lacombar@gmail.com>

I recently got bitten in the ass when pressing Ctrl-C and lost all my current
configuration changes. This patch captures SIGINT and allows the user to save
any changes.

Some code refactoring was made in order to handle the exit behavior.
Signed-off-by: NDavidlohr Bueso <dave@gnu.org>
Signed-off-by: NArnaud Lacombe <lacombar@gmail.com>

__enabled_<sym-name> are only generated on visible or selected entries, do not
reflect the purpose of its introduction.

Fix this by always generating these entries for named symbol.
Reported-by: NRabin Vincent <rabin@rab.in>
Signed-off-by: NArnaud Lacombe <lacombar@gmail.com>

Cc: Srinivas Kandagatla <srinivas.kandagatla@st.com>
Signed-off-by: NArnaud Lacombe <lacombar@gmail.com>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

Help text under choice menu is never displayed because it does not have
symbol name associated with it, however many kconfigs have help text
under choice, assuming that it will be displayed when user selects help.
for example in Kconfig if we have:
choice
        prompt "Choice"
        ---help---
           HELP TEXT ...

config A
        bool "A"

config B
        bool "B"

endchoice

Without this patch "HELP TEXT" is not displayed when user selects help
option when "Choice" is highlighted from menuconfig or xconfig or
gconfig.

This patch changes the logic in menu_get_ext_help to display help for
cases which dont have symbol names like choice.
Signed-off-by: NSrinivas Kandagatla <srinivas.kandagatla@st.com>
Reviewed-by: NStuart Menefy <stuart.menefy@st.com>
Reviewed-by: NArnaud Lacombe <lacombar@gmail.com>
Cc: stable@kernel.org
Signed-off-by: NMichal Marek <mmarek@suse.cz>

After commit 54168578, nohelp_text' is no
longer referenced, nuke it.
Signed-off-by: NArnaud Lacombe <lacombar@gmail.com>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

The two targets `localmodconfig' and `localyesconfig' only differs from the
sed(1) ran on the result of `streamline_config.pl' to convert symbols set to
`modules' to `yes'. This conversion can be made directly from the perl script,
and thus avoid duplicating the command to generate the configuration.

Cc: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: NArnaud Lacombe <lacombar@gmail.com>

The options introduced are --localmodconfig (default) and --localyesconfig.
They match the Makefile target behavior.

Cc: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: NArnaud Lacombe <lacombar@gmail.com>

Cc: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: NArnaud Lacombe <lacombar@gmail.com>

Replace the config_is_*() macros with a variant that allows for grepping
for usage of CONFIG_* options in the code. Usage:

  if (IS_ENABLED(CONFIG_NUMA))

or

  #if IS_ENABLED(CONFIG_NUMA)

The IS_ENABLED() macro evaluates to 1 if the argument is set (to either 'y'
or 'm'), IS_BUILTIN() tests if the option is 'y' and IS_MODULE() test if
the option is 'm'. Only boolean and tristate options are supported.
Reviewed-by: NArnaud Lacombe <lacombar@gmail.com>
Acked-by: NRandy Dunlap <rdunlap@xenotime.net>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

Give the user an opportunity to fix the error or save the configuration
under a different path.
Reported-by: NHiromu Yakura <hiromu1996@gmail.com>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

The specialized printer for headers (espectially autoconf.h) is missing
fixup code for S_HEX symbol's "0x" prefix. As long as kconfig does not
warn for such missing prefix, this code is needed. Fix this.

In the same time, fix some nits in `header_print_symbol()'.

Cc: Randy Dunlap <rdunlap@xenotime.net>
Cc: Mauro Carvalho Chehab <mchehab@infradead.org>
Broken-by: NArnaud Lacombe <lacombar@gmail.com>
Reported-by: NRandy Dunlap <rdunlap@xenotime.net>
Reported-by: NMauro Carvalho Chehab <mchehab@infradead.org>
Acked-by: NMauro Carvalho Chehab <mchehab@redhat.com>
Acked-by: NRandy Dunlap <rdunlap@xenotime.net>
Signed-off-by: NArnaud Lacombe <lacombar@gmail.com>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

After the test

	if (!submenu || ...)
		continue;

the variable `submenu' can _not_ be NULL, so do not test for this
situation.

Cc: Nir Tzachar <nir.tzachar@gmail.com>
Signed-off-by: NArnaud Lacombe <lacombar@gmail.com>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

nconf does not check the validity of the current menu when help is
requested (with either <F2>, '?' or 'h'). This leads to a NULL pointer
dereference when an empty menu is encountered.

The following reduced testcase exposes the problem:

config DEP
        bool

menu "FOO"

config BAR
        bool "BAR"
        depends on DEP

endmenu

Issue will happen when entering menu "FOO" and requesting help.

nconf is the only front-end which do not filter the validity of the
current menu. Such filter can not really happen beforehand as other key
which does not deals with the current menu might be entered by the user,
so just bails out earlier if we encounter an invalid menu.

Cc: Nir Tzachar <nir.tzachar@gmail.com>
Cc: Andrej Gelenberg <andrej.gelenberg@udo.edu>
Reported-by: NAndrej Gelenberg <andrej.gelenberg@udo.edu>
Signed-off-by: NArnaud Lacombe <lacombar@gmail.com>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

nconf is the only front-end which does not use this helper, but prefer
to copy/paste the code. The test wrt. menu validity added in this
version of the code is bogus anyway as an invalid menu will get
dereferenced a few line below by calling menu_get_prompt().

For now, convert nconf to use menu_get_ext_help(), as do every other
front-end. We will deals with menu validity checks properly in a
separate commit.

Cc: Nir Tzachar <nir.tzachar@gmail.com>
Cc: Andrej Gelenberg <andrej.gelenberg@udo.edu>
Signed-off-by: NArnaud Lacombe <lacombar@gmail.com>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

I am seeing Wunused-but-set warning while make nconfig.  Looks like
active_menu is not used. Removing it fixes the warning.
Signed-off-by: NRaghavendra D Prabhu <rprabhu@wnohang.net>
Acked-by: NWANG Cong <xiyou.wangcong@gmail.com>
Signed-off-by: NMichal Marek <mmarek@suse.cz>