* [Buildroot] menuconfig / legacy bug
@ 2013-11-12 21:43 Thomas Petazzoni
2013-11-13 8:40 ` Thomas De Schampheleire
0 siblings, 1 reply; 6+ messages in thread
From: Thomas Petazzoni @ 2013-11-12 21:43 UTC (permalink / raw)
To: buildroot
Hello,
Samuel noticed this today: it seems like we have a bug in menuconfig
that is annoying for the legacy thing.
Scenario to reproduce the bug:
* Checkout commit 38dbd33b91adb4543e922e644b09430fa3836068 (which is
right before the removal of the BR2_PACKAGE_QT_JAVASCRIPTCORE)
option.
* Create a configuration that has BR2_PACKAGE_QT_JAVASCRIPTCORE set.
* Checkout the latest master (or any other commit that has the
BR2_PACKAGE_QT_JAVASCRIPTCORE removal commit)
* Run make menuconfig, it shows that Legacy is detected, which is OK.
* Now, go in the Legacy menu, and try to disable the Qt Javascript
Core option. It doesn't disable it and instead menuconfig "jumps" at
the beginning of the long comment that explains the legacy thing.
The following patch workarounds the problem by removing the problematic
lines of the comment. Basically, all lines that contains only "signs"
and not actual text seem to confuse menuconfig for some reason.
Ideas?
Thomas
diff --git a/Config.in.legacy b/Config.in.legacy
index 587afe1..d4bf7a0 100644
--- a/Config.in.legacy
+++ b/Config.in.legacy
@@ -68,20 +68,17 @@ comment "Legacy detected: check the content of the menu below"
menu "Legacy config options"
if BR2_LEGACY
-comment "----------------------------------------------------"
comment "Your old configuration uses legacy options that no "
comment "longer exist in buildroot, as indicated in the menu "
comment "below. As long as these options stay selected, or in"
comment "case of string options are non-empty, the build "
comment "will fail. "
-comment "* "
comment "Where possible, an automatic conversion from old to "
comment "new symbols has been performed. Before making any "
comment "change in this legacy menu, make sure to exit the "
comment "configuration editor a first time and save the "
comment "configuration. Otherwise, the automatic conversion "
comment "of symbols will be lost. "
-comment "* "
comment "After this initial save, reopen the configuration "
comment "editor, inspect the options selected below, read "
comment "their help texts, and verify/update the new "
@@ -90,12 +87,10 @@ comment "menus. When everything is ok, you can disable the "
comment "legacy options in the menu below. Once you have "
comment "disabled all legacy options, this text will "
comment "disappear and you will be able to start the build. "
-comment "* "
comment "Note: at some point in the future, the oldest legacy"
comment "options will be removed, and configuration files "
comment "that still have those options set, will fail to "
comment "build, or run, in unpredictable ways. "
-comment "----------------------------------------------------"
endif
###############################################################################
--
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com
^ permalink raw reply related [flat|nested] 6+ messages in thread
* [Buildroot] menuconfig / legacy bug
2013-11-12 21:43 [Buildroot] menuconfig / legacy bug Thomas Petazzoni
@ 2013-11-13 8:40 ` Thomas De Schampheleire
2013-11-13 8:42 ` Peter Korsgaard
2013-11-13 8:43 ` Yann E. MORIN
0 siblings, 2 replies; 6+ messages in thread
From: Thomas De Schampheleire @ 2013-11-13 8:40 UTC (permalink / raw)
To: buildroot
Hi Thomas,
On Tue, Nov 12, 2013 at 10:43 PM, Thomas Petazzoni
<thomas.petazzoni@free-electrons.com> wrote:
> Hello,
>
> Samuel noticed this today: it seems like we have a bug in menuconfig
> that is annoying for the legacy thing.
>
> Scenario to reproduce the bug:
>
> * Checkout commit 38dbd33b91adb4543e922e644b09430fa3836068 (which is
> right before the removal of the BR2_PACKAGE_QT_JAVASCRIPTCORE)
> option.
>
> * Create a configuration that has BR2_PACKAGE_QT_JAVASCRIPTCORE set.
>
> * Checkout the latest master (or any other commit that has the
> BR2_PACKAGE_QT_JAVASCRIPTCORE removal commit)
>
> * Run make menuconfig, it shows that Legacy is detected, which is OK.
>
> * Now, go in the Legacy menu, and try to disable the Qt Javascript
> Core option. It doesn't disable it and instead menuconfig "jumps" at
> the beginning of the long comment that explains the legacy thing.
>
> The following patch workarounds the problem by removing the problematic
> lines of the comment. Basically, all lines that contains only "signs"
> and not actual text seem to confuse menuconfig for some reason.
>
> Ideas?
I cannot reproduce this problem on my end. I followed the exact steps
as you described.
This is on Linux Mint 15, not sure which packages would be relevant
here. Ncurses is:
lib32ncurses5-dev: 5.9-10ubuntu4
Best regards,
Thomas
^ permalink raw reply [flat|nested] 6+ messages in thread
* [Buildroot] menuconfig / legacy bug
2013-11-13 8:40 ` Thomas De Schampheleire
@ 2013-11-13 8:42 ` Peter Korsgaard
2013-11-13 8:43 ` Yann E. MORIN
1 sibling, 0 replies; 6+ messages in thread
From: Peter Korsgaard @ 2013-11-13 8:42 UTC (permalink / raw)
To: buildroot
>>>>> "Thomas" == Thomas De Schampheleire <patrickdepinguin@gmail.com> writes:
Hi,
> I cannot reproduce this problem on my end. I followed the exact steps
> as you described.
> This is on Linux Mint 15, not sure which packages would be relevant
> here. Ncurses is:
> lib32ncurses5-dev: 5.9-10ubuntu4
You need to ensure that you haven't scrolled too far down. The extra
comment lines have to be visible on the screen for the bug to occur.
--
Bye, Peter Korsgaard
^ permalink raw reply [flat|nested] 6+ messages in thread
* [Buildroot] menuconfig / legacy bug
2013-11-13 8:40 ` Thomas De Schampheleire
2013-11-13 8:42 ` Peter Korsgaard
@ 2013-11-13 8:43 ` Yann E. MORIN
2013-11-13 9:05 ` Thomas De Schampheleire
1 sibling, 1 reply; 6+ messages in thread
From: Yann E. MORIN @ 2013-11-13 8:43 UTC (permalink / raw)
To: buildroot
Thomas, All,
On Wednesday 13 November 2013 09:40:53 Thomas De Schampheleire wrote:
> Hi Thomas,
>
> On Tue, Nov 12, 2013 at 10:43 PM, Thomas Petazzoni
> <thomas.petazzoni@free-electrons.com> wrote:
> > Hello,
> >
> > Samuel noticed this today: it seems like we have a bug in menuconfig
> > that is annoying for the legacy thing.
> >
> > Scenario to reproduce the bug:
> >
> > * Checkout commit 38dbd33b91adb4543e922e644b09430fa3836068 (which is
> > right before the removal of the BR2_PACKAGE_QT_JAVASCRIPTCORE)
> > option.
> >
> > * Create a configuration that has BR2_PACKAGE_QT_JAVASCRIPTCORE set.
> >
> > * Checkout the latest master (or any other commit that has the
> > BR2_PACKAGE_QT_JAVASCRIPTCORE removal commit)
> >
> > * Run make menuconfig, it shows that Legacy is detected, which is OK.
> >
> > * Now, go in the Legacy menu, and try to disable the Qt Javascript
> > Core option. It doesn't disable it and instead menuconfig "jumps" at
> > the beginning of the long comment that explains the legacy thing.
> >
> > The following patch workarounds the problem by removing the problematic
> > lines of the comment. Basically, all lines that contains only "signs"
> > and not actual text seem to confuse menuconfig for some reason.
> >
> > Ideas?
>
> I cannot reproduce this problem on my end. I followed the exact steps
> as you described.
> This is on Linux Mint 15, not sure which packages would be relevant
> here. Ncurses is:
> lib32ncurses5-dev: 5.9-10ubuntu4
You have to use 'space' to deselect the option. Using 'n' works as expected.
I have a (quick) fix I need to further test tonight, since it may have some
weird side-effects...
Regards,
Yann E. MORIN.
--
.-----------------.--------------------.------------------.--------------------.
| Yann E. MORIN | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +0/33 662376056 | Software Designer | \ / CAMPAIGN | ^ |
| --==< O_o >==-- '------------.-------: X AGAINST | /e\ There is no |
| http://ymorin.is-a-geek.org/ | (*_*) | / \ HTML MAIL | """ conspiracy. |
'------------------------------'-------'------------------'--------------------'
^ permalink raw reply [flat|nested] 6+ messages in thread
* [Buildroot] menuconfig / legacy bug
2013-11-13 8:43 ` Yann E. MORIN
@ 2013-11-13 9:05 ` Thomas De Schampheleire
2013-11-13 21:54 ` Yann E. MORIN
0 siblings, 1 reply; 6+ messages in thread
From: Thomas De Schampheleire @ 2013-11-13 9:05 UTC (permalink / raw)
To: buildroot
Hi Yann, all,
On Wed, Nov 13, 2013 at 9:43 AM, Yann E. MORIN <yann.morin.1998@free.fr> wrote:
> Thomas, All,
>
> On Wednesday 13 November 2013 09:40:53 Thomas De Schampheleire wrote:
>> Hi Thomas,
>>
>> On Tue, Nov 12, 2013 at 10:43 PM, Thomas Petazzoni
>> <thomas.petazzoni@free-electrons.com> wrote:
>> > Hello,
>> >
>> > Samuel noticed this today: it seems like we have a bug in menuconfig
>> > that is annoying for the legacy thing.
>> >
>> > Scenario to reproduce the bug:
>> >
>> > * Checkout commit 38dbd33b91adb4543e922e644b09430fa3836068 (which is
>> > right before the removal of the BR2_PACKAGE_QT_JAVASCRIPTCORE)
>> > option.
>> >
>> > * Create a configuration that has BR2_PACKAGE_QT_JAVASCRIPTCORE set.
>> >
>> > * Checkout the latest master (or any other commit that has the
>> > BR2_PACKAGE_QT_JAVASCRIPTCORE removal commit)
>> >
>> > * Run make menuconfig, it shows that Legacy is detected, which is OK.
>> >
>> > * Now, go in the Legacy menu, and try to disable the Qt Javascript
>> > Core option. It doesn't disable it and instead menuconfig "jumps" at
>> > the beginning of the long comment that explains the legacy thing.
>> >
>> > The following patch workarounds the problem by removing the problematic
>> > lines of the comment. Basically, all lines that contains only "signs"
>> > and not actual text seem to confuse menuconfig for some reason.
>> >
>> > Ideas?
>>
>> I cannot reproduce this problem on my end. I followed the exact steps
>> as you described.
>> This is on Linux Mint 15, not sure which packages would be relevant
>> here. Ncurses is:
>> lib32ncurses5-dev: 5.9-10ubuntu4
>
> You have to use 'space' to deselect the option. Using 'n' works as expected.
>
Confirmed.
Typing 'space' simply cycles between visible comment lines _not_
containing any letter (number fails too).
Adding a letter on each non-letter line is a workaround:
diff --git a/Config.in.legacy b/Config.in.legacy
--- a/Config.in.legacy
+++ b/Config.in.legacy
@@ -68,20 +68,20 @@ comment "Legacy detected: check the cont
menu "Legacy config options"
if BR2_LEGACY
-comment "----------------------------------------------------"
+comment "---------------------------------------------------o"
comment "Your old configuration uses legacy options that no "
comment "longer exist in buildroot, as indicated in the menu "
comment "below. As long as these options stay selected, or in"
comment "case of string options are non-empty, the build "
comment "will fail. "
-comment "* "
+comment "* o"
comment "Where possible, an automatic conversion from old to "
comment "new symbols has been performed. Before making any "
comment "change in this legacy menu, make sure to exit the "
comment "configuration editor a first time and save the "
comment "configuration. Otherwise, the automatic conversion "
comment "of symbols will be lost. "
-comment "* "
+comment "* o"
comment "After this initial save, reopen the configuration "
comment "editor, inspect the options selected below, read "
comment "their help texts, and verify/update the new "
@@ -90,12 +90,12 @@ comment "menus. When everything is ok, y
comment "legacy options in the menu below. Once you have "
comment "disabled all legacy options, this text will "
comment "disappear and you will be able to start the build. "
-comment "* "
+comment "* o"
comment "Note: at some point in the future, the oldest legacy"
comment "options will be removed, and configuration files "
comment "that still have those options set, will fail to "
comment "build, or run, in unpredictable ways. "
-comment "----------------------------------------------------"
+comment "---------------------------------------------------o"
endif
###############################################################################
> I have a (quick) fix I need to further test tonight, since it may have some
> weird side-effects...
Could you elaborate a bit, or maybe show the patch?
Best regards,
Thomas
^ permalink raw reply [flat|nested] 6+ messages in thread
* [Buildroot] menuconfig / legacy bug
2013-11-13 9:05 ` Thomas De Schampheleire
@ 2013-11-13 21:54 ` Yann E. MORIN
0 siblings, 0 replies; 6+ messages in thread
From: Yann E. MORIN @ 2013-11-13 21:54 UTC (permalink / raw)
To: buildroot
Thomas, All,
On 2013-11-13 10:05 +0100, Thomas De Schampheleire spake thusly:
> On Wed, Nov 13, 2013 at 9:43 AM, Yann E. MORIN <yann.morin.1998@free.fr> wrote:
> > I have a (quick) fix I need to further test tonight, since it may have some
> > weird side-effects...
>
> Could you elaborate a bit, or maybe show the patch?
Now posted to the list. ;-)
Regards,
Yann E. MORIN.
--
.-----------------.--------------------.------------------.--------------------.
| Yann E. MORIN | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software Designer | \ / CAMPAIGN | ___ |
| +33 223 225 172 `------------.-------: X AGAINST | \e/ There is no |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL | v conspiracy. |
'------------------------------^-------^------------------^--------------------'
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2013-11-13 21:54 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2013-11-12 21:43 [Buildroot] menuconfig / legacy bug Thomas Petazzoni
2013-11-13 8:40 ` Thomas De Schampheleire
2013-11-13 8:42 ` Peter Korsgaard
2013-11-13 8:43 ` Yann E. MORIN
2013-11-13 9:05 ` Thomas De Schampheleire
2013-11-13 21:54 ` Yann E. MORIN
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.