Processed: severity of 812196 is important

[Debian Bug Tracking System]

Processing commands for cont...@bugs.debian.org:

> severity 812196 important
Bug #812196 [src:linux] linux-image-4.4.0-trunk-amd64: Laptop reboot after 
suspend system
Severity set to 'important' from 'grave'
> thanks
Stopping processing here.

Please contact me if you need assistance.
-- 
812196: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=812196
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems

Bug#812196: linux-image-4.4.0-trunk-amd64: Laptop reboot after suspend system

[Alexandre]

Package: src:linux
Version: 4.4-1~exp1
Severity: grave
Justification: renders package unusable

Dear Maintainer,

After suspend my laptop (hibernation) i have this message in syslog :

Jan 21 12:16:53 debian dnsmasq[986]: no servers found in /etc/resolv.conf, will
retry
Jan 21 12:16:53 debian gnome-session[863]: Window manager warning: Failed to
read EDID of output eDP32: Aucun fichier ou dossier de ce type
Jan 21 12:16:53 debian gnome-session[863]: Window manager warning: Failed to
set power save mode for output eDP32: Permission non accordée
Jan 21 12:16:54 debian kernel: [ 4939.143250] brcmfmac:
brcmf_cfg80211_reg_notifier: not a ISO3166 code
Jan 21 12:16:54 debian kernel: [ 4939.143265] cfg80211: World regulatory domain
updated:
Jan 21 12:16:54 debian kernel: [ 4939.143271] cfg80211:  DFS Master region:
unset
Jan 21 12:16:54 debian kernel: [ 4939.143276] cfg80211:   (start_freq -
end_freq @ bandwidth), (max_antenna_gain, max_eirp), (dfs_cac_time)
Jan 21 12:16:54 debian kernel: [ 4939.143286] cfg80211:   (2402000 KHz -
2472000 KHz @ 4 KHz), (N/A, 2000 mBm), (N/A)
Jan 21 12:16:54 debian kernel: [ 4939.143293] cfg80211:   (2457000 KHz -
2482000 KHz @ 4 KHz), (N/A, 2000 mBm), (N/A)
Jan 21 12:16:54 debian kernel: [ 4939.143299] cfg80211:   (2474000 KHz -
2494000 KHz @ 2 KHz), (N/A, 2000 mBm), (N/A)
Jan 21 12:16:54 debian kernel: [ 4939.143308] cfg80211:   (517 KHz -
525 KHz @ 8 KHz, 16 KHz AUTO), (N/A, 2000 mBm), (N/A)
Jan 21 12:16:54 debian kernel: [ 4939.143316] cfg80211:   (525 KHz -
533 KHz @ 8 KHz, 16 KHz AUTO), (N/A, 2000 mBm), (0 s)
Jan 21 12:16:54 debian kernel: [ 4939.143324] cfg80211:   (549 KHz -
573 KHz @ 16 KHz), (N/A, 2000 mBm), (0 s)
Jan 21 12:16:54 debian kernel: [ 4939.143330] cfg80211:   (5735000 KHz -
5835000 KHz @ 8 KHz), (N/A, 2000 mBm), (N/A)
Jan 21 12:16:54 debian kernel: [ 4939.143337] cfg80211:   (5724 KHz -
6372 KHz @ 216 KHz), (N/A, 0 mBm), (N/A)
Jan 21 12:16:54 debian NetworkManager[707]:   Failed to
GDBus.Error:fi.w1.wpa_supplicant1.NotConnected: This interface is not
connected: disconnect.
Jan 21 12:16:54 debian dbus[712]: [system] Activating via systemd: service
name='org.freedesktop.nm_dispatcher' unit='dbus-org.freedesktop.nm-
dispatcher.service'
Jan 21 12:16:54 debian gnome-session[863]: Gjs-Message: JS LOG: An active
wireless connection, in infrastructure mode, involves no access point?
Jan 21 12:16:54 debian systemd[1]: Starting Network Manager Script Dispatcher
Service...
Jan 21 12:16:54 debian evolution.desktop[2511]: evolution-shell-Message:
Network disconnected.  Forced offline.
Jan 21 12:16:54 debian dbus[712]: [system] Successfully activated service
'org.freedesktop.nm_dispatcher'
Jan 21 12:16:54 debian systemd[1]: Started Network Manager Script Dispatcher
Service.
Jan 21 12:16:54 debian nm-dispatcher: Dispatching action 'down' for wlp58s0
Jan 21 12:16:58 debian org.freedesktop.Telepathy.ConnectionManager.salut[1388]:
tp-glib-Message: Exiting
Jan 21 12:16:58 debian systemd[1]: Reached target Sleep.
Jan 21 12:16:58 debian systemd[1]: Starting Suspend...
Jan 21 12:16:58 debian systemd-sleep[4293]: Suspending system...

When i try to resume my laptop reboot with this message in syslog :

Jan 21 13:41:51 debian kernel: [0.657877] PM: Hibernation image not present
or could not be loaded.
Jan 21 13:41:51 debian kernel: [0.658559] Freeing unused kernel memory:
1296K (81b0b000 - 81c4f000)
Jan 21 13:41:51 debian kernel: [0.658560] Write protecting the kernel read-
only data: 10240k
Jan 21 13:41:51 debian kernel: [0.659407] Freeing unused kernel memory:
468K (88000178b000 - 88000180)
Jan 21 13:41:51 debian kernel: [0.661337] Freeing unused kernel memory:
1296K (880001abc000 - 880001c0)
Jan 21 13:41:51 debian kernel: [0.661340] [ cut here
]
Jan 21 13:41:51 debian kernel: [0.661343] WARNING: CPU: 2 PID: 1 at /build
/linux-tEELBQ/linux-4.4/arch/x86/mm/dump_pagetables.c:225
note_page+0x5e1/0x790()
Jan 21 13:41:51 debian kernel: [0.661344] x86/mm: Found insecure W+X
mapping at address 8805f000/0x8805f000
Jan 21 13:41:51 debian kernel: [0.661345] Modules linked in:
Jan 21 13:41:51 debian kernel: [0.661347] CPU: 2 PID: 1 Comm: swapper/0 Not
tainted 4.4.0-trunk-amd64 #1 Debian 4.4-1~exp1
Jan 21 13:41:51 debian kernel: [0.661348] Hardware name: Dell Inc. XPS 13
9350/0PWNCR, BIOS 1.0.0 09/10/2015
Jan 21 13:41:51 debian kernel: [0.661348]  
5b9bc4bd 812e5fb9 8802b6c4fde0
Jan 21 13:41:51 debian kernel: [0.661350]  810762ad
8802b6c4fed0 8802b6c4fe38 0004
Jan 21 13:41:51 debian kernel: [0.661351]  
8802b6c4fed0 8107633c 817f1538
Jan 21 13:41:51 debian kernel: [0.661353] Call Trace:
Jan 21 13:41:51 debian kernel: [0.661356]  [] ?
dump_stack+0x40/0x57
Jan 21 13:41:51 debian 

Bug#810663: Include Device Tree model in reportbug script

[Roger Shimizu]

On Mon, Jan 11, 2016 at 2:23 PM, Martin Michlmayr  wrote:
> Package: linux
> Version: 4.3.3-5
> Severity: wishlist
> Tags: patch
>
> It would be nice to include the Device Tree model in the reportbug
> output.  On DT based platforms, /proc/cpuinfo only includes quite
> generic information.
>
> Please note that the strange "echo ... $(cat ..)" construct is
> intentional.  'cat /proc/device-tree/model' leads to a strange
> character at the end because there's no newline and using echo
> gets rid of it.
>
> diff --git a/debian/templates/image.plain.bug/include-model 
> b/debian/templates/image.plain.bug/include-model
> index 60a7112..9c6aedd 100644
> --- a/debian/templates/image.plain.bug/include-model
> +++ b/debian/templates/image.plain.bug/include-model
> @@ -39,6 +39,11 @@ grep_model() {
>  false
>  ;;
>esac
> +
> +  # Device Tree model
> +  if [ -r /proc/device-tree/model ]; then
> +echo "Device Tree model:" $(cat /proc/device-tree/model)
> +  fi
>  }
>
>  add_model() {

Above patch seems only report the model from dts.
How about user uses a modified dts, other than the one ships with the kernel?

So I suggest to add a hash-code of the dts used currently.
Is there any method to find the file location of dts used currently?

Cheers,
Roger

Bug#811351: additional request...

[Andrew Lunn]

> I can add more verbose comments to mainline kernel .dts on how to
> enable serial port, and how to select between rs232/485. Andrew, do
> you want me to resend the current patches, or can it be done with an
> incremental patch?

Either, but incremental is probably easiest.

Andrew

Bug#810663: Include Device Tree model in reportbug script

[Uwe Kleine-König]

Hello Roger,

On Fri, Jan 22, 2016 at 12:11:12AM +0900, Roger Shimizu wrote:
> On Mon, Jan 11, 2016 at 2:23 PM, Martin Michlmayr  wrote:
> > Package: linux
> > Version: 4.3.3-5
> > Severity: wishlist
> > Tags: patch
> >
> > It would be nice to include the Device Tree model in the reportbug
> > output.  On DT based platforms, /proc/cpuinfo only includes quite
> > generic information.
> >
> > Please note that the strange "echo ... $(cat ..)" construct is
> > intentional.  'cat /proc/device-tree/model' leads to a strange
> > character at the end because there's no newline and using echo
> > gets rid of it.
> >
> > diff --git a/debian/templates/image.plain.bug/include-model 
> > b/debian/templates/image.plain.bug/include-model
> > index 60a7112..9c6aedd 100644
> > --- a/debian/templates/image.plain.bug/include-model
> > +++ b/debian/templates/image.plain.bug/include-model
> > @@ -39,6 +39,11 @@ grep_model() {
> >  false
> >  ;;
> >esac
> > +
> > +  # Device Tree model
> > +  if [ -r /proc/device-tree/model ]; then
> > +echo "Device Tree model:" $(cat /proc/device-tree/model)
> > +  fi
> >  }
> >
> >  add_model() {
> 
> Above patch seems only report the model from dts.
> How about user uses a modified dts, other than the one ships with the kernel?
> 
> So I suggest to add a hash-code of the dts used currently.
> Is there any method to find the file location of dts used currently?
The location is not available, because it's put in memory by the
bootloader before jumping into the kernel image. So the best thing you
can do is to extract it from /sys. I don't know off-hand how this works,
but I know who to ask if needed.

Best regards
Uwe

-- 
Pengutronix e.K.   | Uwe Kleine-König|
Industrial Linux Solutions | http://www.pengutronix.de/  |

Processed: [bts-link] source package src:linux

[Debian Bug Tracking System]

Processing commands for cont...@bugs.debian.org:

> #
> # bts-link upstream status pull for source package src:linux
> # see http://lists.debian.org/debian-devel-announce/2006/05/msg1.html
> #
> user bts-link-upstr...@lists.alioth.debian.org
Setting user to bts-link-upstr...@lists.alioth.debian.org (was 
bts-link-de...@lists.alioth.debian.org).
> # remote status report for #599768 (http://bugs.debian.org/599768)
> # Bug title: linux-image-2.6.32-5-686: backlight control fails after resume 
> from suspend (toshiba r700-155)
> #  * http://bugzilla.kernel.org/show_bug.cgi?id=21012
> #  * remote status changed: NEW -> RESOLVED
> #  * remote resolution changed: (?) -> CODE-FIX
> #  * closed upstream
> tags 599768 + fixed-upstream
Bug #599768 [src:linux] linux-image-2.6.32-5-686: backlight control fails after 
resume from suspend (toshiba r700-155)
Added tag(s) fixed-upstream.
> usertags 599768 - status-NEW
Usertags were: status-NEW.
Usertags are now: .
> usertags 599768 + status-RESOLVED resolution-CODE-FIX
There were no usertags set.
Usertags are now: status-RESOLVED resolution-CODE-FIX.
> thanks
Stopping processing here.

Please contact me if you need assistance.
-- 
599768: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=599768
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems

Bug#810663: Include Device Tree model in reportbug script

[Roger Shimizu]

On Fri, Jan 22, 2016 at 1:44 AM, Uwe Kleine-König
 wrote:
> Hello Roger,
>
> On Fri, Jan 22, 2016 at 12:11:12AM +0900, Roger Shimizu wrote:
>> On Mon, Jan 11, 2016 at 2:23 PM, Martin Michlmayr  wrote:
>> > Package: linux
>> > Version: 4.3.3-5
>> > Severity: wishlist
>> > Tags: patch
>> >
>> > It would be nice to include the Device Tree model in the reportbug
>> > output.  On DT based platforms, /proc/cpuinfo only includes quite
>> > generic information.
>> >
>> > Please note that the strange "echo ... $(cat ..)" construct is
>> > intentional.  'cat /proc/device-tree/model' leads to a strange
>> > character at the end because there's no newline and using echo
>> > gets rid of it.
>> >
>> > diff --git a/debian/templates/image.plain.bug/include-model 
>> > b/debian/templates/image.plain.bug/include-model
>> > index 60a7112..9c6aedd 100644
>> > --- a/debian/templates/image.plain.bug/include-model
>> > +++ b/debian/templates/image.plain.bug/include-model
>> > @@ -39,6 +39,11 @@ grep_model() {
>> >  false
>> >  ;;
>> >esac
>> > +
>> > +  # Device Tree model
>> > +  if [ -r /proc/device-tree/model ]; then
>> > +echo "Device Tree model:" $(cat /proc/device-tree/model)
>> > +  fi
>> >  }
>> >
>> >  add_model() {
>>
>> Above patch seems only report the model from dts.
>> How about user uses a modified dts, other than the one ships with the kernel?
>>
>> So I suggest to add a hash-code of the dts used currently.
>> Is there any method to find the file location of dts used currently?
> The location is not available, because it's put in memory by the
> bootloader before jumping into the kernel image. So the best thing you
> can do is to extract it from /sys. I don't know off-hand how this works,
> but I know who to ask if needed.

According to: http://xillybus.com/tutorials/device-tree-zynq-1
the DTS can be extracted by:

scripts/dtc/dtc -I fs -O dts -o ~/effective.dts /proc/device-tree/

I haven't tried this, but if so and we need to add the dependency of
dtc command, then we can append the DTS to bugreport.

Cheers,
Roger

Bug#810663: Include Device Tree model in reportbug script

[Roger Shimizu]

On Fri, Jan 22, 2016 at 1:53 AM, Roger Shimizu  wrote:
> On Fri, Jan 22, 2016 at 1:44 AM, Uwe Kleine-König
>>> Above patch seems only report the model from dts.
>>> How about user uses a modified dts, other than the one ships with the 
>>> kernel?
>>>
>>> So I suggest to add a hash-code of the dts used currently.
>>> Is there any method to find the file location of dts used currently?
>> The location is not available, because it's put in memory by the
>> bootloader before jumping into the kernel image. So the best thing you
>> can do is to extract it from /sys. I don't know off-hand how this works,
>> but I know who to ask if needed.
>
> According to: http://xillybus.com/tutorials/device-tree-zynq-1
> the DTS can be extracted by:
>
> scripts/dtc/dtc -I fs -O dts -o ~/effective.dts /proc/device-tree/
>
> I haven't tried this, but if so and we need to add the dependency of
> dtc command, then we can append the DTS to bugreport.

Can help to try this earlier. So there's the result

# apt install device-tree-compiler
# dtc -I fs /proc/device-tree/

so you can get the full DTS

Cheers,
Roger

Processed: [PATCH initramfs-tools] scripts/functions: Connect panic shell to normal tty device

[Debian Bug Tracking System]

Processing control commands:

> tag -1 patch pending
Bug #512679 [initramfs-tools] initramfs-tools: 'more' pager is broken in 
initramfs shell
Added tag(s) patch and pending.

-- 
512679: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=512679
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems

Bug#512679: [PATCH initramfs-tools] scripts/functions: Connect panic shell to normal tty device

[Ben Hutchings]

Control: tag -1 patch pending

I worked out how to do this without setsid -c.

Ben.
---
Currently the panic shell's controlling tty is /dev/console which is
not fully functional - the shell can't provide job control and more
can't work out the screen size for paging.

Fix this by reading /proc/consoles to find out the underlying tty
device and then connecting the shell to it directly with the aid of
setsid.

Closes: #512679
Signed-off-by: Ben Hutchings 
---
 scripts/functions | 17 -
 1 file changed, 16 insertions(+), 1 deletion(-)

diff --git a/scripts/functions b/scripts/functions
index 531d1bd..67ce319 100644
--- a/scripts/functions
+++ b/scripts/functions
@@ -33,6 +33,8 @@ log_end_msg()
 
 panic()
 {
+   local console rest
+
if command -v chvt >/dev/null 2>&1; then
chvt 1
fi
@@ -48,7 +50,20 @@ panic()
 
run_scripts /scripts/panic
 
-   REASON="$@" PS1='(initramfs) ' /bin/sh -i /dev/console 
2>&1
+   # Try to use setsid, which will enable job control in the shell
+   # and paging in more
+   if command -v setsid >/dev/null 2>&1; then
+   read console rest /dev/${console} 1>&0 2>&1"
+   else
+   REASON="$@" PS1='(initramfs) ' sh -i /dev/console 2>&1
+   fi
 }
 
 maybe_break()


signature.asc
Description: Digital signature

Processed: tagging 810154, tagging 767471, tagging 804071

[Debian Bug Tracking System]

Processing commands for cont...@bugs.debian.org:

> tags 810154 + patch
Bug #810154 [initramfs-tools] remove busybox hook, leave responsibility to 
busybox package
Added tag(s) patch.
> tags 767471 + patch
Bug #767471 [initramfs-tools] [initramfs-tools] / formatted as ext3 but mounted 
as ext4 not fsck-able
Added tag(s) patch.
> tags 804071 + patch
Bug #804071 [initramfs-tools] initramfs-tools: does not respect the periodic 
fsck interval for ext4
Added tag(s) patch.
> thanks
Stopping processing here.

Please contact me if you need assistance.
-- 
767471: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=767471
804071: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=804071
810154: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=810154
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems

Bug#804071: [PATCH initramfs-tools 1/2] hook-functions: Include any modular RTC drivers in the initramfs

[Ben Hutchings]

The RTC is needed for periodic filesystem checks to work.  Debian
currently builds-in all RTC drivers though.

Signed-off-by: Ben Hutchings 
---
 hook-functions | 5 -
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/hook-functions b/hook-functions
index 5da0cde..817d5fc 100644
--- a/hook-functions
+++ b/hook-functions
@@ -405,7 +405,7 @@ dep_add_modules_mount()
block_dev_mod_add "$dev_node"
 
# sys walk some important device classes
-   for class in gpio phy regulator; do
+   for class in gpio phy regulator rtc; do
for device in /sys/class/$class/*; do
device="$(readlink -f "$device")" \
&& sys_walk_mod_add "$device"
@@ -502,6 +502,9 @@ auto_add_modules()
copy_modules_dir kernel/drivers/pinctrl
copy_modules_dir kernel/drivers/regulator
copy_modules_dir kernel/drivers/usb/phy
+
+   # Needed for periodic fsck
+   copy_modules_dir kernel/drivers/rtc
;;
net)
copy_modules_dir kernel/drivers/net \



signature.asc
Description: Digital signature

[PATCH initramfs-tools 0/4] Improve early keyboard driver loading

[Ben Hutchings]

The first possible breakpoint (break=top) comes before we start
udev or load manually configurd modules.  Thus only built-in
keyboard drivers can be used.  We have a bodge to load a list of
known USB controller drivers and usbhid, but this still isn't
sufficient because we also need hid_generic or a more specific
HID driver.

These changes make break=top work for me when using a USB keyboard
in QEMU.  They also suppress the warning messages which currently
appear when starting a shell (unless using break=top and no quiet
parameter).

Ben.

Ben Hutchings (4):
  scripts/functions: Explicitly load keyboard drivers only at break=top
  scripts/functions: Use more sensible modprobe options at break=top
  scripts/functions: Load more USB host controllers drivers on break=top
  scripts/functions: Load HID drivers at break=top

 scripts/functions | 25 +
 1 file changed, 17 insertions(+), 8 deletions(-)



signature.asc
Description: Digital signature

[PATCH initramfs-tools 1/4] scripts/functions: Explicitly load keyboard drivers only at break=top

[Ben Hutchings]

If we open a shell when udev is already running, any necessary drivers
should already have been loaded.

Signed-off-by: Ben Hutchings 
---
 scripts/functions | 19 +++
 1 file changed, 11 insertions(+), 8 deletions(-)

diff --git a/scripts/functions b/scripts/functions
index 930a12c..2156ce5 100644
--- a/scripts/functions
+++ b/scripts/functions
@@ -45,14 +45,6 @@ panic()
reboot
exit  # in case reboot fails, force kernel panic
fi
-   modprobe -v i8042 || true
-   modprobe -v atkbd || true
-   modprobe -v ehci-pci || true
-   modprobe -v ehci-orion || true
-   modprobe -v ehci-hcd || true
-   modprobe -v uhci-hcd || true
-   modprobe -v ohci-hcd || true
-   modprobe -v usbhid || true
 
run_scripts /scripts/panic
 
@@ -63,6 +55,17 @@ maybe_break()
 {
case ",$break," in
*,$1,*)
+   if [ "$1" = "top" ]; then
+   # udev is not yet running, so load keyboard drivers
+   modprobe -v i8042 || true
+   modprobe -v atkbd || true
+   modprobe -v ehci-pci || true
+   modprobe -v ehci-orion || true
+   modprobe -v ehci-hcd || true
+   modprobe -v uhci-hcd || true
+   modprobe -v ohci-hcd || true
+   modprobe -v usbhid || true
+   fi
panic "Spawning shell within the initramfs"
;;
esac



signature.asc
Description: Digital signature

[PATCH initramfs-tools 2/4] scripts/functions: Use more sensible modprobe options at break=top

[Ben Hutchings]

Use -a to load everything in one command.

Use -q rather than -v when quiet=y.

Signed-off-by: Ben Hutchings 
---
 scripts/functions | 15 +++
 1 file changed, 7 insertions(+), 8 deletions(-)

diff --git a/scripts/functions b/scripts/functions
index 2156ce5..f549ba4 100644
--- a/scripts/functions
+++ b/scripts/functions
@@ -57,14 +57,13 @@ maybe_break()
*,$1,*)
if [ "$1" = "top" ]; then
# udev is not yet running, so load keyboard drivers
-   modprobe -v i8042 || true
-   modprobe -v atkbd || true
-   modprobe -v ehci-pci || true
-   modprobe -v ehci-orion || true
-   modprobe -v ehci-hcd || true
-   modprobe -v uhci-hcd || true
-   modprobe -v ohci-hcd || true
-   modprobe -v usbhid || true
+   if [ "${quiet}" = "y" ]; then
+   opts="-q"
+   else
+   opts="-v"
+   fi
+   modprobe ${opts} -a i8042 atkbd ehci-pci ehci-orion \
+ehci-hcd uhci-hcd ohci-hcd usbhid
fi
panic "Spawning shell within the initramfs"
;;



signature.asc
Description: Digital signature

[PATCH initramfs-tools 3/4] scripts/functions: Load more USB host controllers drivers on break=top

[Ben Hutchings]

Signed-off-by: Ben Hutchings 
---
 scripts/functions | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/scripts/functions b/scripts/functions
index f549ba4..2fc3ab2 100644
--- a/scripts/functions
+++ b/scripts/functions
@@ -63,7 +63,8 @@ maybe_break()
opts="-v"
fi
modprobe ${opts} -a i8042 atkbd ehci-pci ehci-orion \
-ehci-hcd uhci-hcd ohci-hcd usbhid
+ehci-hcd ohci-hcd ohci-pci uhci-hcd usbhid 
xhci \
+xhci-pci xhci-hcd
fi
panic "Spawning shell within the initramfs"
;;



signature.asc
Description: Digital signature

Bug#767471: [PATCH initramfs-tools 0/4] Fix filesystem type confusion

[Ben Hutchings]

This series changes probing of filesystem type at initramfs build and
boot time to more closely match the old behaviour of initscripts.

Ben.

Ben Hutchings (4):
  Assume blkid is always installed, and remove fallbacks
  hooks/fsck: Use get_fstype rather than blkid to probe filesystem type
  hooks/fsck: Always use get_fstype to probe type of root filesystem
  Pass filesystem type from /etc/fstab to fsck when checking /usr

 hooks/fsck| 28 +++-
 hooks/resume  |  5 ++---
 scripts/functions | 17 +++--
 scripts/local |  4 ++--
 4 files changed, 22 insertions(+), 32 deletions(-)



signature.asc
Description: Digital signature

Bug#767471: [PATCH initramfs-tools 4/4] Pass filesystem type from /etc/fstab to fsck when checking /usr

[Ben Hutchings]

Unlike root, the types of all other filesystems in /etc/fstab have
historically been honoured and we should continue to do so.

Signed-off-by: Ben Hutchings 
---
 scripts/functions | 6 +-
 scripts/local | 4 ++--
 2 files changed, 7 insertions(+), 3 deletions(-)

diff --git a/scripts/functions b/scripts/functions
index 2c36ed2..990fe6c 100644
--- a/scripts/functions
+++ b/scripts/functions
@@ -306,17 +306,21 @@ resolve_device() {
 # Check a file system.
 # $1=device
 # $2=mountpoint (for diagnostics only)
+# $3=type (may be "auto")
 _checkfs_once()
 {
DEV="$1"
NAME="$2"
+   TYPE="$3"
if [ "$NAME" = "/" ] ; then
NAME="root"
fi
FSCK_LOGFILE=/run/initramfs/fsck.log
FSCK_STAMPFILE=/run/initramfs/fsck-${NAME#/}
 
-   TYPE=$(get_fstype "$1")
+   if [ "${TYPE}" = "auto" ]; then
+   TYPE="$(get_fstype "${DEV}")"
+   fi
 
FSCKCODE=0
if ! command -v fsck >/dev/null 2>&1; then
diff --git a/scripts/local b/scripts/local
index af56e66..3558f7a 100644
--- a/scripts/local
+++ b/scripts/local
@@ -136,7 +136,7 @@ local_mount_root()
# FIXME This has no error checking
modprobe ${FSTYPE}
 
-   checkfs ${ROOT} root
+   checkfs ${ROOT} root "${FSTYPE}"
 
# FIXME This has no error checking
# Mount root
@@ -166,7 +166,7 @@ local_mount_fs()
modprobe "${MNT_TYPE}"
 
if [ "$MNT_PASS" != 0 ]; then
-   checkfs "$MNT_FSNAME" "$MNT_DIR"
+   checkfs "$MNT_FSNAME" "$MNT_DIR" "${MNT_TYPE}"
fi
 
# FIXME This has no error checking


signature.asc
Description: Digital signature

Bug#767471: [PATCH initramfs-tools 2/4] hooks/fsck: Use get_fstype rather than blkid to probe filesystem type

[Ben Hutchings]

This makes the behaviour consistent with boot time.

Signed-off-by: Ben Hutchings 
---
 hooks/fsck | 10 --
 1 file changed, 4 insertions(+), 6 deletions(-)

diff --git a/hooks/fsck b/hooks/fsck
index ad0fbc5..486a04f 100755
--- a/hooks/fsck
+++ b/hooks/fsck
@@ -58,11 +58,9 @@ get_fsck_type_fstab () {
case "$MNT_TYPE" in
auto)
MNT_FSNAME=$(resolve_device "$MNT_FSNAME")
-   MNT_TYPE=$(blkid -o value -s TYPE 
"${MNT_FSNAME}")
-   if [ -z "${MNT_TYPE}" ]; then
-   MNT_TYPE="auto"
-   fi
-   echo "$MNT_TYPE"
+   alias fstype="/usr/lib/klibc/bin/fstype"
+   get_fstype "${MNT_FSNAME}"
+   unalias fstype
;;
*)
echo "$MNT_TYPE"
@@ -101,7 +99,7 @@ copy_exec /sbin/fsck
 copy_exec /sbin/logsave
 
 for type in $fsck_types; do
-   if [ "$type" = 'auto' ] ; then
+   if [ "$type" = "unknown" ] ; then
echo "Warning: couldn't identify filesystem type for fsck hook, 
ignoring."
continue
fi



signature.asc
Description: Digital signature

Bug#767471: [PATCH initramfs-tools 1/4] Assume blkid is always installed, and remove fallbacks

[Ben Hutchings]

udev has used blkid since version 142, so we can rely on it being
present at build and boot time.

Signed-off-by: Ben Hutchings 
---
 hooks/fsck|  6 ++
 hooks/resume  |  5 ++---
 scripts/functions | 11 ++-
 3 files changed, 6 insertions(+), 16 deletions(-)

diff --git a/hooks/fsck b/hooks/fsck
index 4472f6a..ad0fbc5 100755
--- a/hooks/fsck
+++ b/hooks/fsck
@@ -57,10 +57,8 @@ get_fsck_type_fstab () {
if [ "$1" = "$MNT_DIR" ] && [ "$MNT_PASS" != 0 ]; then
case "$MNT_TYPE" in
auto)
-   if command -v blkid >/dev/null 2>&1 ; then
-   MNT_FSNAME=$(resolve_device 
"$MNT_FSNAME")
-   MNT_TYPE=$(blkid -o value -s TYPE 
"${MNT_FSNAME}")
-   fi
+   MNT_FSNAME=$(resolve_device "$MNT_FSNAME")
+   MNT_TYPE=$(blkid -o value -s TYPE 
"${MNT_FSNAME}")
if [ -z "${MNT_TYPE}" ]; then
MNT_TYPE="auto"
fi
diff --git a/hooks/resume b/hooks/resume
index 96e2ffd..0d33977 100755
--- a/hooks/resume
+++ b/hooks/resume
@@ -18,8 +18,7 @@ esac
 # First check if a location is set and is a valid swap partition
 test -r /etc/initramfs-tools/conf.d/resume \
&& . /etc/initramfs-tools/conf.d/resume
-if [ -n "$RESUME" ] && command -v blkid >/dev/null 2>&1 \
-   && blkid -p -n swap $RESUME >/dev/null 2>&1; then
+if [ -n "$RESUME" ] && blkid -p -n swap $RESUME >/dev/null 2>&1; then
# As mkinitramfs copies the config file nothing to do.
exit 0
 fi
@@ -42,7 +41,7 @@ chrooted() {
 
 # Try to autodetect the RESUME partition, using biggest swap?
 RESUME=$(grep ^/dev/ /proc/swaps | sort -rk3 | head -n 1 | cut -d " " -f 1)
-if [ -n "$RESUME" ] && command -v blkid >/dev/null 2>&1; then
+if [ -n "$RESUME" ]; then
UUID=$(blkid -s UUID -o value "$RESUME" || true)
if [ -n "$UUID" ]; then
RESUME="UUID=$UUID"
diff --git a/scripts/functions b/scripts/functions
index 930a12c..2c36ed2 100644
--- a/scripts/functions
+++ b/scripts/functions
@@ -143,10 +143,8 @@ get_fstype ()
# but fstype is more robust
FSTYPE="unknown"
eval $(fstype "${FS}" 2> /dev/null)
-   if [ "$FSTYPE" = "unknown" ] &&  command -v blkid >/dev/null 2>&1 ; then
+   if [ "$FSTYPE" = "unknown" ]; then
FSTYPE=$(blkid -o value -s TYPE "${FS}")
-   elif [ "$FSTYPE" = "unknown" ] && [ -x /lib/udev/vol_id ]; then
-   FSTYPE=$(/lib/udev/vol_id -t "${FS}" 2> /dev/null)
fi
RET=$?
 
@@ -299,12 +297,7 @@ resolve_device() {
 
case "$DEV" in
LABEL=* | UUID=* | PARTLABEL=* | PARTUUID=*)
-   if command -v blkid >/dev/null 2>&1; then
-   DEV="$(blkid -l -t "$DEV" -o device)" || return 1
-   else
-   log_warning_msg "blkid not present, so cannot resolve 
$DEV"
-   return 1
-   fi
+   DEV="$(blkid -l -t "$DEV" -o device)" || return 1
;;
esac
[ -e "$DEV" ] && echo "$DEV"



signature.asc
Description: Digital signature

Bug#767471: [PATCH initramfs-tools 3/4] hooks/fsck: Always use get_fstype to probe type of root filesystem

[Ben Hutchings]

We can't trust the type in /etc/fstab because it is not used when
mounting the root filesystem.

Closes: #767471
Signed-off-by: Ben Hutchings 
---
 hooks/fsck | 22 ++
 1 file changed, 10 insertions(+), 12 deletions(-)

diff --git a/hooks/fsck b/hooks/fsck
index 486a04f..86d0dde 100755
--- a/hooks/fsck
+++ b/hooks/fsck
@@ -55,18 +55,16 @@ get_fsck_type_fstab () {
 
# Not found by default.
if [ "$1" = "$MNT_DIR" ] && [ "$MNT_PASS" != 0 ]; then
-   case "$MNT_TYPE" in
-   auto)
-   MNT_FSNAME=$(resolve_device "$MNT_FSNAME")
-   alias fstype="/usr/lib/klibc/bin/fstype"
-   get_fstype "${MNT_FSNAME}"
-   unalias fstype
-   ;;
-   *)
-   echo "$MNT_TYPE"
-   ;;
-   esac
-
+   # Ignore filesystem type for /, as it is not available and
+   # therefore never used at boot time
+   if [ "${MNT_DIR}" = "/" ] || [ "${MNT_TYPE}" = "auto" ]; then
+   MNT_FSNAME="$(resolve_device "${MNT_FSNAME}")"
+   alias fstype="/usr/lib/klibc/bin/fstype"
+   get_fstype "${MNT_FSNAME}"
+   unalias fstype
+   else
+   echo "${MNT_TYPE}"
+   fi
fi
 }
 



signature.asc
Description: Digital signature

[PATCH initramfs-tools 4/4] scripts/functions: Load HID drivers at break=top

[Ben Hutchings]

Load modules for all devices found on the HID bus.

Signed-off-by: Ben Hutchings 
---
 scripts/functions | 6 ++
 1 file changed, 6 insertions(+)

diff --git a/scripts/functions b/scripts/functions
index 2fc3ab2..a6687c1 100644
--- a/scripts/functions
+++ b/scripts/functions
@@ -65,6 +65,12 @@ maybe_break()
modprobe ${opts} -a i8042 atkbd ehci-pci ehci-orion \
 ehci-hcd ohci-hcd ohci-pci uhci-hcd usbhid 
xhci \
 xhci-pci xhci-hcd
+   sleep 2
+   for modalias in /sys/bus/hid/devices/*/modalias; do
+   if [ -f "${modalias}" ]; then
+   modprobe ${opts} -b "$(cat ${modalias})"
+   fi
+   done
fi
panic "Spawning shell within the initramfs"
;;


signature.asc
Description: Digital signature

Bug#810154: [PATCH initramfs-tools 2/4] debian/control: Recommend the right busybox packages for the target distribution

[Ben Hutchings]

Ubuntu's busybox and busybox-static are no use in an initramfs, so
instead of listing all variants of busybox, check for Ubuntu
derivatives at build time.

Signed-off-by: Ben Hutchings 
---
 debian/control | 2 +-
 debian/rules   | 9 +
 2 files changed, 10 insertions(+), 1 deletion(-)

diff --git a/debian/control b/debian/control
index f31dd05..9263e1c 100644
--- a/debian/control
+++ b/debian/control
@@ -24,7 +24,7 @@ Description: generic modular initramfs generator (automation)
 Package: initramfs-tools-core
 Architecture: all
 Multi-Arch: foreign
-Recommends: busybox (>= 1:1.01-3) | busybox-initramfs | busybox-static
+Recommends: ${busybox:Recommends}
 Depends: klibc-utils (>= 2.0.4-7~), cpio, kmod | module-init-tools, udev, 
${misc:Depends}
 Suggests: bash-completion
 Breaks: initramfs-tools (<< 0.121~)
diff --git a/debian/rules b/debian/rules
index e6e920c..ea2ff3b 100755
--- a/debian/rules
+++ b/debian/rules
@@ -2,3 +2,12 @@
 
 %:
dh --with bash-completion $@
+
+# On Debian we can use either busybox or busybox-static, but on Ubuntu
+# and derivatives only busybox-initramfs will work.
+BUSYBOX_PACKAGES := $(shell if dpkg-vendor --derives-from ubuntu; then echo 
busybox-initramfs; else echo busybox busybox-static; fi)
+BUSYBOX_MIN_VERSION := 1:1.01-3
+
+override_dh_gencontrol:
+   echo >> debian/initramfs-tools-core.substvars 
"busybox:Recommends=$(wordlist 2,100,$(BUSYBOX_PACKAGES:%=| % (>= 
$(BUSYBOX_MIN_VERSION"
+   dh_gencontrol



signature.asc
Description: Digital signature

Bug#810154: [PATCH initramfs-tools 3/4] Remove busybox hook in favour of busybox's own

[Ben Hutchings]

- Set BUSYBOXDIR empty initially and let busybox's conf hook override it
- Fail if BUSYBOX=y and BUSYBOXDIR remains empty after running conf hooks
- Bump the minimum version to the first version that has the conf hook
- Update the automatic selection of klibc utilities to check whether
  BUSYBOXDIR is empty

Closes: #810154
Signed-off-by: Ben Hutchings 
---
 debian/initramfs-tools-core.dirs|  1 +
 debian/initramfs-tools-core.install |  1 -
 debian/rules|  9 -
 hooks/busybox   | 34 --
 hooks/klibc |  3 +--
 mkinitramfs | 10 --
 6 files changed, 18 insertions(+), 40 deletions(-)
 delete mode 100755 hooks/busybox

diff --git a/debian/initramfs-tools-core.dirs b/debian/initramfs-tools-core.dirs
index bcb978b..21f605c 100644
--- a/debian/initramfs-tools-core.dirs
+++ b/debian/initramfs-tools-core.dirs
@@ -10,6 +10,7 @@ etc/initramfs-tools/scripts/nfs-top
 etc/initramfs-tools/scripts/panic
 etc/initramfs-tools/hooks
 etc/initramfs-tools/conf.d
+usr/sbin
 usr/share/initramfs-tools/conf.d
 usr/share/initramfs-tools/conf-hooks.d
 usr/share/initramfs-tools/modules.d
diff --git a/debian/initramfs-tools-core.install 
b/debian/initramfs-tools-core.install
index 91d0378..2b57045 100644
--- a/debian/initramfs-tools-core.install
+++ b/debian/initramfs-tools-core.install
@@ -1,5 +1,4 @@
 lsinitramfsusr/bin
-mkinitramfsusr/sbin
 init   usr/share/initramfs-tools
 scriptsusr/share/initramfs-tools
 conf/initramfs.confetc/initramfs-tools
diff --git a/debian/rules b/debian/rules
index ea2ff3b..d630547 100755
--- a/debian/rules
+++ b/debian/rules
@@ -6,8 +6,15 @@
 # On Debian we can use either busybox or busybox-static, but on Ubuntu
 # and derivatives only busybox-initramfs will work.
 BUSYBOX_PACKAGES := $(shell if dpkg-vendor --derives-from ubuntu; then echo 
busybox-initramfs; else echo busybox busybox-static; fi)
-BUSYBOX_MIN_VERSION := 1:1.01-3
+BUSYBOX_MIN_VERSION := 1:1.22.0-17~
 
 override_dh_gencontrol:
echo >> debian/initramfs-tools-core.substvars 
"busybox:Recommends=$(wordlist 2,100,$(BUSYBOX_PACKAGES:%=| % (>= 
$(BUSYBOX_MIN_VERSION"
dh_gencontrol
+
+override_dh_install:
+   sed -e 's,@BUSYBOX_PACKAGES@,$(wordlist 2,100,$(BUSYBOX_PACKAGES:%=or 
%)),' \
+   -e 's/@BUSYBOX_MIN_VERSION@/$(BUSYBOX_MIN_VERSION)/' \
+   mkinitramfs > debian/initramfs-tools-core/usr/sbin/mkinitramfs
+   chmod 755 debian/initramfs-tools-core/usr/sbin/mkinitramfs
+   dh_install
diff --git a/hooks/busybox b/hooks/busybox
deleted file mode 100755
index c01d06d..000
--- a/hooks/busybox
+++ /dev/null
@@ -1,34 +0,0 @@
-#!/bin/sh
-
-PREREQ=""
-
-prereqs()
-{
-   echo "$PREREQ"
-}
-
-case $1 in
-# get pre-requisites
-prereqs)
-   prereqs
-   exit 0
-   ;;
-esac
-
-# busybox
-if [ "${BUSYBOX}" = "n" ]; then
-   exit 0
-fi
-if ! [ -e ${BUSYBOXDIR}/busybox ]; then
-   if [ "${BUSYBOX}" = "y" ]; then
-   echo >&2 "E: busybox is required but not installed"
-   exit 1
-   else
-   exit 0
-   fi
-fi
-. /usr/share/initramfs-tools/hook-functions
-rm -f ${DESTDIR}/bin/sh
-rm -f ${DESTDIR}/bin/busybox
-copy_exec ${BUSYBOXDIR}/busybox /bin/busybox
-ln -s busybox ${DESTDIR}/bin/sh
diff --git a/hooks/klibc b/hooks/klibc
index 2e753ed..d985036 100755
--- a/hooks/klibc
+++ b/hooks/klibc
@@ -19,8 +19,7 @@ esac
 cp -pnL /usr/lib/klibc/bin/* ${DESTDIR}/bin
 cp -pL /lib/klibc-*.so ${DESTDIR}/lib
 rm -f ${DESTDIR}/bin/kinit* ${DESTDIR}/bin/zcat
-if [ "${BUSYBOX}" = "n" ] || \
-   { [ "${BUSYBOX}" != "y" ] && ! [ -e ${BUSYBOXDIR}/busybox ]; }; then
+if [ "${BUSYBOX}" = "n" ] || [ -z "${BUSYBOXDIR}" ]; then
if [ -e ${DESTDIR}/bin/sh.shared ]; then
# Some platforms build a shared klibc/sh:
mv ${DESTDIR}/bin/sh.shared ${DESTDIR}/bin/sh
diff --git a/mkinitramfs b/mkinitramfs
index fac0049..5cc6608 100755
--- a/mkinitramfs
+++ b/mkinitramfs
@@ -7,8 +7,8 @@ export PATH='/usr/bin:/sbin:/bin'
 keep="n"
 CONFDIR="/etc/initramfs-tools"
 verbose="n"
-test -e /bin/busybox && BUSYBOXDIR=/bin
-test -e /usr/lib/initramfs-tools/bin/busybox && 
BUSYBOXDIR=/usr/lib/initramfs-tools/bin
+# Will be updated by busybox's conf hook, if present
+BUSYBOXDIR=
 export BUSYBOXDIR
 
 OPTIONS=`getopt -o c:d:ko:r:v -n "$0" -- "$@"`
@@ -90,6 +90,12 @@ for i in /usr/share/initramfs-tools/conf-hooks.d/*; do
fi
 done
 
+# Check busybox dependency
+if [ "${BUSYBOX}" = "y" ] && [ -z "${BUSYBOXDIR}" ]; then
+   echo >&2 "E: @BUSYBOX_PACKAGES@, version @BUSYBOX_MIN_VERSION@ or 
later, is required but not installed"
+   exit 1
+fi
+
 if [ -n "${UMASK:-}" ]; then
umask "${UMASK}"
 fi



signature.asc
Description: Digital signature

Processed: tagging 639876

[Debian Bug Tracking System]

Processing commands for cont...@bugs.debian.org:

> tags 639876 + patch pending
Bug #639876 [initramfs-tools] MODULES=dep does not include input drivers
Bug #701656 [initramfs-tools] Include current input drivers when MODULES=dep, 
to support rescue shell
Bug #772578 [initramfs-tools] [initramfs-tools] Missing keyboard modules 
i2c_designware_* in initramfs
Added tag(s) patch and pending.
Added tag(s) pending and patch.
Added tag(s) pending and patch.
> thanks
Stopping processing here.

Please contact me if you need assistance.
-- 
639876: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=639876
701656: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=701656
772578: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=772578
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems

Processed: tagging 810154

[Debian Bug Tracking System]

Processing commands for cont...@bugs.debian.org:

> tags 810154 + pending
Bug #810154 [initramfs-tools] remove busybox hook, leave responsibility to 
busybox package
Ignoring request to alter tags of bug #810154 to the same tags previously set
> thanks
Stopping processing here.

Please contact me if you need assistance.
-- 
810154: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=810154
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems

Bug#639876: [PATCH initramfs-tools] hook-functions: Include drivers for all keyboards when MODULES=dep

[Ben Hutchings]

Additional modular drivers (e.g. ohci_pci, hid) may be needed for
entering a disk encryption passphrase or for the panic shell.

Closes: #639876
Signed-off-by: Ben Hutchings 
---
 hook-functions | 11 +++
 1 file changed, 11 insertions(+)

diff --git a/hook-functions b/hook-functions
index 5da0cde..42cee54 100644
--- a/hook-functions
+++ b/hook-functions
@@ -418,6 +418,17 @@ dep_add_modules_mount()
add_loaded_modules 'phy[-_]*'
add_loaded_modules 'pinctrl[-_]*'
 
+   # Sys walk keyboards.  We identify keyboards as input devices
+   # that can generate at least key events 1-31; udev has the
+   # same heuristic.  Note that the format of the bitmap
+   # properties depends on the word size of the process reading
+   # the uevent file!
+   for device in /sys/class/input/input*; do
+   if grep -qs "^KEY=.*fff[ef]$" "${device}/uevent"; then
+   sys_walk_mod_add "$(readlink -f "$device")"
+   fi
+   done
+
# catch old-style IDE
if [ -e /sys/bus/ide/devices/ ]; then
modules="$modules ide-gd_mod ide-cd"


signature.asc
Description: Digital signature

Bug#804071: [PATCH initramfs-tools 2/2] hooks/fsck: Stop setting e2fsck broken_system_clock option

[Ben Hutchings]

e2fsck no longer prompts whether to fix mount or write times that are
up to 24 hours ahead of the current time; it always does so.  Update
the old NEWS entry about this, and add a versioned Breaks to ensure we
get the new version.

Closes: #804071
Signed-off-by: Ben Hutchings 
---
 debian/NEWS| 2 ++
 debian/control | 2 +-
 hooks/fsck | 9 -
 3 files changed, 3 insertions(+), 10 deletions(-)

diff --git a/debian/NEWS b/debian/NEWS
index 52a8f9b..e862c61 100644
--- a/debian/NEWS
+++ b/debian/NEWS
@@ -30,6 +30,8 @@ initramfs-tools (0.119) unstable; urgency=medium
 following lines in /etc/e2fsck.conf:
 [options]
 broken_system_clock=1
+[As of e2fsprogs version 1.42.13 this message is informational, and
+ no configuration change is required.]
 
  -- Ben Hutchings   Mon, 13 Apr 2015 01:00:21 +0100
 
diff --git a/debian/control b/debian/control
index a943598..94862c2 100644
--- a/debian/control
+++ b/debian/control
@@ -15,7 +15,7 @@ Depends: initramfs-tools-core (= ${binary:Version}), 
linux-base, ${misc:Depends}
 Suggests: bash-completion
 Provides: linux-initramfs-tool
 Conflicts: linux-initramfs-tool, usplash (<< 0.5.50)
-Breaks: cryptsetup (<< 2:1.6.6-4~), elilo (<< 3.12-3.1~), lilo (<< 22.8-8.2~), 
s390-tools (<< 1.8.3-2~), console-setup (<< 1.72), systemd-sysv (<< 186), lvm2 
(<< 2.02.111-2.1~), initscripts (<< 2.88dsf-59.3~), upstart
+Breaks: cryptsetup (<< 2:1.6.6-4~), elilo (<< 3.12-3.1~), lilo (<< 22.8-8.2~), 
s390-tools (<< 1.8.3-2~), console-setup (<< 1.72), systemd-sysv (<< 186), lvm2 
(<< 2.02.111-2.1~), initscripts (<< 2.88dsf-59.3~), upstart, e2fsprogs (<< 
1.42.13)
 Description: generic modular initramfs generator (automation)
  This package builds a bootable initramfs for Linux kernel packages.  The
  initramfs is loaded along with the kernel and is responsible for
diff --git a/hooks/fsck b/hooks/fsck
index 0cebf94..4472f6a 100755
--- a/hooks/fsck
+++ b/hooks/fsck
@@ -102,15 +102,6 @@ fi
 copy_exec /sbin/fsck
 copy_exec /sbin/logsave
 
-# e2fsck may fail or warn if the system time is not yet correct, which
-# will happen if the RTC driver is modular or the RTC is set to local
-# time.  Disable this behaviour.  (No other fsck does this, apparently.)
-mkdir -p "${DESTDIR}/etc"
-cat >"${DESTDIR}/etc/e2fsck.conf" <

signature.asc
Description: Digital signature

Bug#810154: [PATCH initramfs-tools 1/4] debian/control: Remove redundant dependency on klibc-utils|busybox

[Ben Hutchings]

As we already have a dependency on a newer version of klibc-utils
(for run-init), the last dependency has no effect.

Signed-off-by: Ben Hutchings 
---
 debian/control | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/debian/control b/debian/control
index a943598..f31dd05 100644
--- a/debian/control
+++ b/debian/control
@@ -25,7 +25,7 @@ Package: initramfs-tools-core
 Architecture: all
 Multi-Arch: foreign
 Recommends: busybox (>= 1:1.01-3) | busybox-initramfs | busybox-static
-Depends: klibc-utils (>= 2.0.4-7~), cpio, kmod | module-init-tools, udev, 
${misc:Depends}, klibc-utils (>= 2.0.4-1.2~) | busybox (>= 1:1.01-3) | 
busybox-initramfs | busybox-static
+Depends: klibc-utils (>= 2.0.4-7~), cpio, kmod | module-init-tools, udev, 
${misc:Depends}
 Suggests: bash-completion
 Breaks: initramfs-tools (<< 0.121~)
 Replaces: initramfs-tools (<< 0.121~)



signature.asc
Description: Digital signature

Bug#810154: [PATCH initramfs-tools 0/4] Changes to busybox integration

[Ben Hutchings]

This series removes the busybox hook script and definition of
BUSYBOXDIR from initramfs-tools, leaving busybox itself responsible
for these.

busybox has installed its own hook script for some time, but will
now need to define BUSYBOXDIR in a conf hook script.  That change
is pending in git but not yet uploaded.

This updates the recommendation of busybox accordingly.

Ben.

Ben Hutchings (4):
  debian/control: Remove redundant dependency on klibc-utils|busybox
  debian/control: Recommend the right busybox packages for the target
distribution
  Remove busybox hook in favour of busybox's own
  Update TODO for removal of busybox hook

 debian/TODO |  2 +-
 debian/control  |  4 ++--
 debian/initramfs-tools-core.dirs|  1 +
 debian/initramfs-tools-core.install |  1 -
 debian/rules| 16 
 hooks/busybox   | 34 --
 hooks/klibc |  3 +--
 mkinitramfs | 10 --
 8 files changed, 29 insertions(+), 42 deletions(-)
 delete mode 100755 hooks/busybox



signature.asc
Description: Digital signature

Bug#810154: [PATCH initramfs-tools 4/4] Update TODO for removal of busybox hook

[Ben Hutchings]

Signed-off-by: Ben Hutchings 
---
 debian/TODO | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/debian/TODO b/debian/TODO
index dee4900..9617b0c 100644
--- a/debian/TODO
+++ b/debian/TODO
@@ -12,4 +12,4 @@ TODO
 
  o root loop support.
 
- o Eliminate ?klibc?, busybox (-> glibc).
+ o Eliminate klibc hook?


signature.asc
Description: Digital signature

Bug#588770: [PATCH initramfs-tools 1/2] hook-functions: Clean up split between dep_add_modules and dep_add_modules_mount

[Ben Hutchings]

Much of the code in dep_add_modules is unrelated to any particular mount
point and should only be run once.

Signed-off-by: Ben Hutchings 
---
 hook-functions | 18 ++
 1 file changed, 10 insertions(+), 8 deletions(-)

diff --git a/hook-functions b/hook-functions
index bbb4d7e..6790d76 100644
--- a/hook-functions
+++ b/hook-functions
@@ -318,7 +318,6 @@ add_loaded_modules()
 dep_add_modules_mount()
 {
local dir dev_node FSTYPE
-   local modules=
 
dir="$1"
 
@@ -400,9 +399,18 @@ dep_add_modules_mount()
fi
 
# Add filesystem
-   modules="$modules ${FSTYPE}"
+   manual_add_modules ${FSTYPE}
 
block_dev_mod_add "$dev_node"
+}
+
+dep_add_modules()
+{
+   local device
+   local modules=
+
+   dep_add_modules_mount /
+   dep_add_modules_mount /usr
 
# sys walk some important device classes
for class in gpio phy regulator rtc; do
@@ -462,12 +470,6 @@ dep_add_modules_mount()
manual_add_modules $modules
 }
 
-dep_add_modules()
-{
-   dep_add_modules_mount /
-   dep_add_modules_mount /usr
-}
-
 # The modules "most" classes added per default to the initramfs
 auto_add_modules()
 {



signature.asc
Description: Digital signature

Bug#811496: initramfs-tools: obsolete conffile /etc/bash_completion.d/initramfs-tools

[Ben Hutchings]

Control: reassign -1 bash-completion 1:2.1-4.2

On Tue, 19 Jan 2016 14:58:29 +0100 Jakub Wilk  wrote:
> Package: initramfs-tools
> Version: 0.121
> User: debian...@lists.debian.org
> Usertags: adequate obsolete-conffile
> 
> The package left obsolete conffile after upgrade:
> /etc/bash_completion.d/initramfs-tools
> 
> 
> 
> This bug was brought to you by adequate:
> https://packages.debian.org/unstable/main/adequate
> 

This was introduced by a change in dh_bash-completion and should be
fixed there.

Ben.

-- 
Ben Hutchings
Quantity is no substitute for quality, but it's the only one we've got.

signature.asc
Description: This is a digitally signed message part

Processed: Re: initramfs-tools: obsolete conffile /etc/bash_completion.d/initramfs-tools

[Debian Bug Tracking System]

Processing control commands:

> reassign -1 bash-completion 1:2.1-4.2
Bug #811496 [initramfs-tools] initramfs-tools: obsolete conffile 
/etc/bash_completion.d/initramfs-tools
Bug reassigned from package 'initramfs-tools' to 'bash-completion'.
No longer marked as found in versions initramfs-tools/0.121.
Ignoring request to alter fixed versions of bug #811496 to the same values 
previously set
Bug #811496 [bash-completion] initramfs-tools: obsolete conffile 
/etc/bash_completion.d/initramfs-tools
Marked as found in versions bash-completion/1:2.1-4.2.

-- 
811496: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=811496
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems

Processed: Re: dpkg: warning: unable to delete old directory '/etc/initramfs-tools/conf.d': Directory not empty

[Debian Bug Tracking System]

Processing control commands:

> tag -1 unreproducible moreinfo
Bug #809206 [initramfs-tools] dpkg: warning: unable to delete old directory 
'/etc/initramfs-tools/conf.d': Directory not empty
Added tag(s) moreinfo and unreproducible.

-- 
809206: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=809206
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems

Processed: Re: initramfs-tools: mkinitramfs uses ldd, which is insecure and generates core dumps

[Debian Bug Tracking System]

Processing control commands:

> tag -1 moreinfo
Bug #720735 [initramfs-tools] initramfs-tools: Use static check for library 
dependencies instead of ldd
Added tag(s) moreinfo.

-- 
720735: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=720735
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems

Bug#720735: initramfs-tools: mkinitramfs uses ldd, which is insecure and generates core dumps

[Ben Hutchings]

Control: tag -1 moreinfo

On Sun, 25 Aug 2013 00:37:53 +0200 Vincent Lefevre 
wrote:
> Package: initramfs-tools
> Version: 0.113
> Severity: important
> Tags: security
> 
> I've noticed that when running update-initramfs, a core dump was
> generated in the current directory, which is in itself a first bug.
> 
> After looking at this problem with strace, I saw that this came from:
> 
>   /usr/bin/ldd /lib/firmware/cis/PCMLM28.cis
[...]

In version 0.121~rc1 the copy_exec function has been split up and we
should now only be running ldd when copying executables.

These executables are being copied and used in the initramfs so they
are already trusted.  So I don't think there's any security reason to
move away from using ldd.

Do you think there is still a bug to fix here, or can this be closed?

Ben.

-- 
Ben Hutchings
Quantity is no substitute for quality, but it's the only one we've got


signature.asc
Description: This is a digitally signed message part

Bug#768913: [PATCH initramfs-tools] kernel/postinst.d: Make test for update-initramfs consistent with invocation

[Ben Hutchings]

Control: tag -1 patch pending
---
We test whether /usr/sbin/update-initramfs exists, but then invoke it
as 'update-initramfs'.  Change the test to look in $PATH.

Closes: #768913
Signed-off-by: Ben Hutchings 
---
 kernel/postinst.d/initramfs-tools | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/kernel/postinst.d/initramfs-tools 
b/kernel/postinst.d/initramfs-tools
index 50f244a..0d463f4 100755
--- a/kernel/postinst.d/initramfs-tools
+++ b/kernel/postinst.d/initramfs-tools
@@ -3,7 +3,7 @@
 version="$1"
 bootopt=""
 
-[ -x /usr/sbin/update-initramfs ] || exit 0
+command -v update-initramfs >/dev/null 2>&1 || exit 0
 
 # passing the kernel version is required
 if [ -z "${version}" ]; then


signature.asc
Description: Digital signature

Processed: [PATCH initramfs-tools] kernel/postinst.d: Make test for update-initramfs consistent with invocation

[Debian Bug Tracking System]

Processing control commands:

> tag -1 patch pending
Bug #768913 [initramfs-tools] /etc/kernel/postinst.d/initramfs-tools: fragile 
PATH dependency
Added tag(s) pending and patch.

-- 
768913: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=768913
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems

Bug#804295: initramfs-tools: doesn't warn/fail when dpkg triggers for update-initramfs don't actually update

[Ben Hutchings]

On Sat, 07 Nov 2015 02:47:27 +0100 Christoph Anton Mitterer 
 wrote:
> Package: initramfs-tools
> Version: 0.120
> Severity: normal
> 
> 
> Hi.
> 
> I've just noted the following:
> Processing triggers for initramfs-tools (0.120) ...
> update-initramfs: /boot/initrd.img-4.2.0-1-amd64 has been altered.
> update-initramfs: Cannot update. Override with -t option.

So it does actually warn, contrary to the bug title.

> Not sure what caused it to think that I've modified it,

mkinitramfs saves md5sums under /var/lib/initramfs-tools.  Perhaps you
replaced the initramfs image with a different version, or perhaps it
was corrupted at some point.

> ... nevertheless.
> such a status message will likely just drown in the flood of log messages
> during any update.

I don't think you can blame the package maintainer if you ignore such
warnings.

> And in such case the initramfs may stay stale, which could cause quite
> some troubles,... from non booting systems up to even security issues.

Kernel updates always run update-initramfs synchronously using the -t
option (via /etc/kernel/postinst.d/initramfs-tools) so I don't see much
of a security issue here.

> I'm not sure whether it would be better to simply let the trigger use -t,
> cause this may be undesired either.

I'm not sure how important the -t option is, now that we have
initramfs-tools-core for users that don't want automatic rebuilds.
Maybe we can get rid of it.

> Would there be a way to give a more interactive warning (e.g. debconf)?
> Or does it seem reasonable in such a case to fail the trigger?

I don't like either of those options.  I think the answer will usually
be that we should go ahead and update the initramfs anyway.

Ben.

-- 
Ben Hutchings
Quantity is no substitute for quality, but it's the only one we've got.

signature.asc
Description: This is a digitally signed message part

Processed: retitle 792910 to Should not explicitly load drivers found through sysfs walk

[Debian Bug Tracking System]

Processing commands for cont...@bugs.debian.org:

> retitle 792910 Should not explicitly load drivers found through sysfs walk
Bug #792910 [initramfs-tools] Module aliases should be resolved at build time
Changed Bug title to 'Should not explicitly load drivers found through sysfs 
walk' from 'Module aliases should be resolved at build time'
> thanks
Stopping processing here.

Please contact me if you need assistance.
-- 
792910: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=792910
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems

Processed: tagging 792910

[Debian Bug Tracking System]

Processing commands for cont...@bugs.debian.org:

> tags 792910 + patch pending
Bug #792910 [initramfs-tools] Should not explicitly load drivers found through 
sysfs walk
Added tag(s) pending and patch.
> thanks
Stopping processing here.

Please contact me if you need assistance.
-- 
792910: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=792910
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems

Processed: [PATCH initramfs-tools 2/2] hook-functions: Include modules for resume partition when MODULES=dep

[Debian Bug Tracking System]

Processing control commands:

> tag -1 patch pending
Bug #588770 [initramfs-tools] dep_add_modules does not consider resume partition
Added tag(s) patch and pending.

-- 
588770: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=588770
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems

Bug#588770: [PATCH initramfs-tools 2/2] hook-functions: Include modules for resume partition when MODULES=dep

[Ben Hutchings]

Control: tag -1 patch pending
---
Closes: #588770
Signed-off-by: Ben Hutchings 
---
 hook-functions | 9 -
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/hook-functions b/hook-functions
index 6790d76..6aef679 100644
--- a/hook-functions
+++ b/hook-functions
@@ -406,12 +406,19 @@ dep_add_modules_mount()
 
 dep_add_modules()
 {
-   local device
+   local device dev_node
local modules=
 
dep_add_modules_mount /
dep_add_modules_mount /usr
 
+   if [ -n "${RESUME}" ]; then
+   dev_node="$(resolve_device "${RESUME}")"
+   if [ -n "${dev_node}" ]; then
+   block_dev_mod_add "${dev_node}"
+   fi
+   fi
+
# sys walk some important device classes
for class in gpio phy regulator rtc; do
for device in /sys/class/$class/*; do


signature.asc
Description: Digital signature

Bug#792910: [PATCH initramfs-tools] hook-functions: Stop force-loading drivers found through sysfs when MODULES=dep

[Ben Hutchings]

Currently we're duplicating the work of udev and generating warnings
for drivers that are always built-in.

Closes: #792910
Signed-off-by: Ben Hutchings 
---
 hook-functions | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/hook-functions b/hook-functions
index 6aef679..60f798c 100644
--- a/hook-functions
+++ b/hook-functions
@@ -228,7 +228,7 @@ sys_walk_mod_add()
if [ -e "${device_path}/modalias" ]; then
modalias=$(cat "${device_path}/modalias")
if [ -n "${modalias}" ]; then
-   force_load "${modalias}"
+   manual_add_modules "${modalias}"
fi
fi
 
@@ -237,7 +237,7 @@ sys_walk_mod_add()
if [ -e "$driver_path" ]; then
module="$(basename $(readlink -f $driver_path))"
if [ -n "${module}" ]; then
-   force_load "${module}"
+   manual_add_modules "${module}"
fi
fi
 


signature.asc
Description: Digital signature

Processed: severity of 636697 is wishlist ...

[Debian Bug Tracking System]

Processing commands for cont...@bugs.debian.org:

> severity 636697 wishlist
Bug #636697 [initramfs-tools] initramfs-tools: no way to include library 
modules for libraries installed in multiarch path
Severity set to 'wishlist' from 'normal'
> retitle 636697 Add a common function to include library modules along with a 
> shared library
Bug #636697 [initramfs-tools] initramfs-tools: no way to include library 
modules for libraries installed in multiarch path
Changed Bug title to 'Add a common function to include library modules along 
with a shared library' from 'initramfs-tools: no way to include library modules 
for libraries installed in multiarch path'
> thanks
Stopping processing here.

Please contact me if you need assistance.
-- 
636697: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=636697
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems

Processed: Re: /etc/kernel/postinst.d/initramfs-tools: fragile PATH dependency

[Debian Bug Tracking System]

Processing control commands:

> clone -1 -2
Bug #768913 [initramfs-tools] /etc/kernel/postinst.d/initramfs-tools: fragile 
PATH dependency
Bug 768913 cloned as bug 812266
> reassign -2 debianutils
Bug #812266 [initramfs-tools] /etc/kernel/postinst.d/initramfs-tools: fragile 
PATH dependency
Bug reassigned from package 'initramfs-tools' to 'debianutils'.
No longer marked as found in versions initramfs-tools/0.109.1 and 
initramfs-tools/0.120.
Ignoring request to alter fixed versions of bug #812266 to the same values 
previously set
> retitle -2 installkernel: Should only run scripts if install dir is /boot
Bug #812266 [debianutils] /etc/kernel/postinst.d/initramfs-tools: fragile PATH 
dependency
Changed Bug title to 'installkernel: Should only run scripts if install dir is 
/boot' from '/etc/kernel/postinst.d/initramfs-tools: fragile PATH dependency'
> severity -1 minor
Bug #768913 [initramfs-tools] /etc/kernel/postinst.d/initramfs-tools: fragile 
PATH dependency
Severity set to 'minor' from 'normal'

-- 
768913: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=768913
812266: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=812266
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems

Bug#768913: /etc/kernel/postinst.d/initramfs-tools: fragile PATH dependency

[Ben Hutchings]

Control: clone -1 -2
Control: reassign -2 debianutils
Control: retitle -2 installkernel: Should only run scripts if install dir is 
/boot
Control: severity -1 minor

On Sun, 09 Nov 2014 21:37:14 -0800 Rob Leslie  wrote:
> Package: initramfs-tools
> Version: 0.109.1
> Severity: normal
> File: /etc/kernel/postinst.d/initramfs-tools
> 
> The script /etc/kernel/postinst.d/initramfs-tools checks for the existence
> of /usr/sbin/update-initramfs using an absolute path, but later invokes it
> without any path. Among other possible dangers, this causes problems when
> /usr/sbin is not in one's PATH.
> 
> I discovered this when trying to build a kernel with kernel-package,
> fakeroot, and zsh (which by default does not include /usr/sbin in PATH for
> ordinary users).
[...]

I acknowledge that the script should be made consistent, but it's not a
practical issue normally.

It looks like you're seeing a bug in the /sbin/installkernel script
(which is also my fault): running 'make install' as non-root, to a
directory other than /boot, still ends up running the post-installation 
scripts.

Ben.

-- 
Ben Hutchings
Quantity is no substitute for quality, but it's the only one we've got.

signature.asc
Description: This is a digitally signed message part

Bug#785526: firmware-linux-nonfree: initramfs trigger only rebuilds initdisk of latest kernel

[Ben Hutchings]

On Sun, 17 May 2015 16:02:08 +0200 Ralf Jung  wrote:
> Hi,
> 
> >> when the package firmware-linux-nonfree is installed, it triggers the hook 
> >> of the initramdisk
> >> package. I conclude that some firmware may be needed by the initramfs 
> >> (which makes sense to me),
> >> and hence it is important to keep the initramfs up-to-date as firmware is 
> >> updated.
> >>
> >> However, only the initramdisk of the latest installed kernel is actually 
> >> rebuilt:
> > [...]
> > 
> > This is controlled by initramfs-tools and it is intentional.  I don't
> > know what the reasoning is, though.
> 
> Interesting, and thanks for reassining.
> In my case, of course, this is all but helpful - I do have a bunch of
> vanilla upstream kernels installed from some recent regression test,
> including some which are way newer than the current Debian kernel (hence
> the weird version number). I am however actually using the latest Debian
> kernel. (I don't reboot that often, but when I do, I simply pick the
> right kernel manually.) Hence the hook doesn't even bother to update the
> initramfs for the kernel I am actually using.

update-initramfs.conf(5) documents how to change this behaviour.

I will consider changing the default, but this would have to be done
only if we also create backups by default.

Ben.
 
-- 
Ben Hutchings
Quantity is no substitute for quality, but it's the only one we've got.

signature.asc
Description: This is a digitally signed message part

Bug#809206: dpkg: warning: unable to delete old directory '/etc/initramfs-tools/conf.d': Directory not empty

[Ben Hutchings]

Control: tag -1 unreproducible moreinfo

On Mon, 28 Dec 2015 18:06:11 +0800 =?utf-8?B?56mN5Li55bC8?= Dan Jacobson 
 wrote:
> Package: initramfs-tools
> Version: 0.121~rc2
> 
> Unpacking initramfs-tools (0.121~rc2) over (0.120) ...
> dpkg: warning: unable to delete old directory '/etc/initramfs-tools/conf.d': 
> Directory not empty
> dpkg: warning: unable to delete old directory '/var/lib/initramfs-tools': 
> Directory not empty
[...]

I can't reproduce this, using either APT or direct invocation of dpkg. 
Please explain how you attempted to upgrade.

Ben.

-- 
Ben Hutchings
Quantity is no substitute for quality, but it's the only one we've got.

signature.asc
Description: This is a digitally signed message part

Bug#810154: [PATCH initramfs-tools 0/4] Changes to busybox integration

[Michael Tokarev]

22.01.2016 01:14, Ben Hutchings wrote:
> This series removes the busybox hook script and definition of
> BUSYBOXDIR from initramfs-tools, leaving busybox itself responsible
> for these.

Oh well.  How many times I talked with Max on IRC, sent patches,
created a git tree for initramfs to pull from..  His answer has
always been the same: no need.  So I gave up, creating an ugly
zzz-busybox which undoes the mess done in initramfs script.

Please note that once the d-i team prevented me from maintaining
busybox, this package remains unmaintained.  So maybe it is a
better idea to remove usage of busybox in initramfs (which this
series actually does).

Thank you Ben!

(And yes, I'm still subscribed to busybox package, for unknown
reason).

/mjt

Bug#809206: dpkg: warning: unable to delete old directory '/etc/initramfs-tools/conf.d': Directory not empty

[積丹尼 Dan Jacobson]

> "BH" == Ben Hutchings  writes:

BH> I can't reproduce this, using either APT or direct invocation of dpkg.
BH> Please explain how you attempted to upgrade.

I always use aptitude full-upgrade.

Anyway on my jidanni2 machine I also find:

$ set /etc/initramfs-tools/conf.d /var/lib/initramfs-tools
$ find $@ -type f|xargs ls -og
-rw-r--r-- 1 70 2010-05-06  /etc/initramfs-tools/conf.d/resume
-rw-r--r-- 1 17 2008-02-12  /etc/initramfs-tools/conf.d/resume^old
-rw-r--r-- 1 75 12-08 08:43 /var/lib/initramfs-tools/4.2.0-1-686-pae
-rw-r--r-- 1 75 01-22 10:39 /var/lib/initramfs-tools/4.3.0-1-686-pae
$ find $@ -type f|xargs more|cat
::
/etc/initramfs-tools/conf.d/resume
::
# RESUME=/dev/hda5
RESUME='UUID=89a43990-ecd9-426f-adf2-529902935af4'
::
/etc/initramfs-tools/conf.d/resume^old
::
RESUME=/dev/hda5
::
/var/lib/initramfs-tools/4.3.0-1-686-pae
::
1dba8e95df7fe33aa6c404433e548e93524fb3ae  /boot/initrd.img-4.3.0-1-686-pae
::
/var/lib/initramfs-tools/4.2.0-1-686-pae
::
a9763ca08e6e8cea06bcd66ff03ddeb201bf793b  /boot/initrd.img-4.2.0-1-686-pae

Bug#810726: Affected by the same bug

[Dominic Harvey]

Hi Kernel team,

I just want to tell you that i am victim of this same bug.

Thank you for working on it.!

[root@dominic-box ~]# cat /proc/version
Linux version 4.3.3-300.fc23.x86_64 (
mockbu...@bkernel01.phx2.fedoraproject.org) (gcc version 5.3.1 20151207
(Red Hat 5.3.1-2) (GCC) ) #1 SMP Tue Jan 5 23:31:01 UTC 2016