Age | Commit message (Collapse) | Author |
|
You are hard pressed to find a true single core x86 chip now,
even those which are a single core typically have hyperthread,
and so we'd want SMP+SCHED_SMT from the smp config frag anyway.
Delete the now redundant include from the existing BSPs too.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
It is on by default for 64bit x86 and only optional if you
are x86 32 bit. Ensure it is on for the latter so we are
aligned with the kernel.org x86 defconfig.
This was listed as a feature, but since HPET has been around
for a while now, we'll not likely be carrying extra patches
for HPET as per the other features.
By sticking it in the shared x86 and x86_64 frags, we can
delete BSP specific mentions of it.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
At the moment, there are no 8250 patches applied, however
if there were, we'd run into issues where they were not
applied at the standard (pre-bsp) level, but instead applied
once per each BSP, duplicating content with unique commit IDs.
Separate out the 8250 enablement Kconfig settings to a cfg
frag, and ensure that the feature [which may someday optionally
contain patches] gets processed independent of the BSPs.
The BSPs which were asking for the feature now just ask for
the enabling config frag in cfg dir.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
This will allow us to delete repeated boiler plate
settings.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
There will be a significant amount of cfg data that is common
across all x86 BSPs. Make a landing ground for it here, similar
to what ARM has, so that we don't duplicate a bunch of stuff
in every x86 BSP file.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Due to historical reasons, the standard.cfg and rt.cfg were
setting a LOG_BUF_SHIFT value that was (at that point in time)
larger than the mainline kernel default value. But time has
moved on, and now it is smaller.
This in turn spawned a features/logbuf/size-normal.cfg fragment
that set a larger value, but time again has moved past this too,
and so the "size-normal" value is once again smaller than the
mainline default (which is CONFIG_LOG_BUF_SHIFT=18).
Get rid of all references; either directly or via reference to
the size-normal thing, and just implicitly take the mainline
default, so we grow with it and don't have to revisit this
in another year or two.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
1/1 [
Author: Bruce Ashfield
Email: bruce.ashfield@windriver.com
Subject: routerstationpro: lib/kobject_uevent: switch to uevent_sock_mutex
Date: Tue, 12 Jun 2012 16:51:14 -0400
-stable commit: 377c2f4aa985f937d77a6bb9b938b4deda2dc282 removed
the old sequence_lock and replaced it with uevent_sock_mutex.
Switching to the new lock to fix the resulting build error.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
]
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Upstream mainline kernel commit f0d1b3c2bcc5de8a1
"net/wanrouter: Deprecate and schedule for removal"
is the justification.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
|
|
1/2 [
Author: Darren Hart
Email: dvhart@linux.intel.com
Subject: pch_uart: Add eg20t_port lock field, avoid recursive spinlocks
Date: Thu, 31 May 2012 00:37:51 -0700
pch_uart_interrupt() takes priv->port.lock which leads to two recursive
spinlock calls if low_latency==1 or CONFIG_PREEMPT_RT_FULL=y (one
otherwise):
pch_uart_interrupt
spin_lock_irqsave(priv->port.lock, flags)
case PCH_UART_IID_RDR_TO (data ready)
handle_rx_to
push_rx
tty_port_tty_get
spin_lock_irqsave(&port->lock, flags) <--- already hold this lock
...
tty_flip_buffer_push
...
flush_to_ldisc
spin_lock_irqsave(&tty->buf.lock)
spin_lock_irqsave(&tty->buf.lock)
disc->ops->receive_buf(tty, char_buf)
n_tty_receive_buf
tty->ops->flush_chars()
uart_flush_chars
uart_start
spin_lock_irqsave(&port->lock) <--- already hold this lock
Avoid this by using a dedicated lock to protect the eg20t_port structure
and IO access to its membase. This is more consistent with the 8250
driver. Ensure priv->lock is always take prior to priv->port.lock when
taken at the same time.
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
CC: Tomoya MORINAGA <tomoya.rohm@gmail.com>
CC: Feng Tang <feng.tang@intel.com>
CC: Alexander Stein <alexander.stein@systec-electronic.com>
CC: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
CC: Alan Cox <alan@linux.intel.com>
CC: linux-serial@vger.kernel.org
]
2/2 [
Author: Darren Hart
Email: dvhart@linux.intel.com
Subject: pch_uart: Correct spin_lock and irqsave usage in pch_console_write
Date: Thu, 31 May 2012 01:21:45 -0700
pch_console_write was calling:
local_irq_save()
spin_lock(a)
spin_lock(b)
...
spin_unlock(b)
spin_unlock(a)
local_irq_restore()
Which trips the rtmutex debug code:
BUG: sleeping function called from invalid context at kernel/rtmutex.c:646
Update the function to use the _nort versions of local_irq* (thanks Peterz for
the suggestion).
local_irq_save_nort()
spin_lock(a)
spin_lock(b)
...
spin_unlock(b)
spin_unlock(a)
local_irq_restore_nort()
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
CC: Tomoya MORINAGA <tomoya.rohm@gmail.com>
CC: Feng Tang <feng.tang@intel.com>
CC: Alexander Stein <alexander.stein@systec-electronic.com>
CC: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
CC: Alan Cox <alan@linux.intel.com>
CC: Steven Rostedt <rostedt@goodmis.org>
CC: Peter Zijlstra <peterz@infradead.org>
CC: linux-serial@vger.kernel.org
]
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
For preempt-rt the fri2 BSP has some board specific patches, so
we need a branch to isolate them from the rest of the boards.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Include the USB CDC_NCM driver as built-in. Comment out i2c features that are
not yet ready for use.
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
|
|
The standard and preempt-rt kernels both support live boot and are expected to
support general use cases. Include support for usb-mass-storage.
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
|
|
Drop support for hardware that isn't present on the FRI2 or that is included by
other features (such as usb-mass-storage).
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
|
|
o Make use of the new USB HCD fragments
o Drop the unused iwlagn driver
o Add the i2c gpio and accelerometer devices
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
|
|
Initial support for Ivy Bridge + Panther Point Intel systems.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Remove the BLK_DEV_SD option already covered by the usb-mass-storage
feature.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Remove the BLK_DEV_SD option already covered by the usb-mass-storage
feature.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Remove the BLK_DEV_SD option already covered by the usb-mass-storage
feature.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Remove features and options already inherited from common-pc-64.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Remove features and options already inherited from common-pc-64.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Remove the BLK_DEV_SR option from common-pc-64.cfg and replace it with
the scsi/cdrom feature.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Add the usb-mass-storage feature and remove the options available
in that feature from common-pc-64.cfg.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Add the usb-mass-storage feature and remove the options available
in that feature from common-pc.cfg.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Refactor bsp/common-pc/sugarbay* to keep hardware-related options in
the base sugarbay.scc while moving the more 'policy'-related options
into the files that include the base sugarbay.scc fragment.
This allows the different kernel types to make use of the basic
machine support without dragging along possibly unwanted options that
fall under the heading of 'policy' for a specific kernel type.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Refactor bsp/common-pc/jasperforest* to keep hardware-related options
in the base jasperforest.scc while moving the more 'policy'-related
options into the files that include the base jasperforest.scc
fragment.
This allows the different kernel types to make use of the basic
machine support without dragging along possibly unwanted options that
fall under the heading of 'policy' for a specific kernel type.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Refactor bsp/common-pc/fishriver* to keep hardware-related options in
the base fishriver.scc while moving the more 'policy'-related options
into the files that include the base fishriver.scc fragment.
This allows the different kernel types to make use of the basic
machine support without dragging along possibly unwanted options that
fall under the heading of 'policy' for a specific kernel type.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Refactor bsp/emenlow/emenlow* to keep hardware-related options in the
base emenlow.scc while moving the more 'policy'-related options into
the files that include the base emenlow.scc fragment.
This allows the different kernel types to make use of the basic
machine support without dragging along possibly unwanted options that
fall under the heading of 'policy' for a specific kernel type.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Refactor bsp/common-pc/crownbay* to keep hardware-related options in
the base crownbay.scc while moving the more 'policy'-related options
into the files that include the base crownbay.scc fragment.
This allows the different kernel types to make use of the basic
machine support without dragging along possibly unwanted options that
fall under the heading of 'policy' for a specific kernel type.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Refactor bsp/common-pc-64/common-pc-64* to keep hardware-related
options in the base common-pc-64.scc while moving the more
'policy'-related options into the files that include the base
common-pc-64.scc fragment.
This allows the different kernel types to make use of the basic
machine support without dragging along possibly unwanted options that
fall under the heading of 'policy' for a specific kernel type.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Refactor bsp/common-pc/common-pc* to keep hardware-related options in
the base common-pc.scc while moving the more 'policy'-related options
into the files that include the base common-pc.scc fragment.
This allows the different kernel types to make use of the basic
machine support without dragging along possibly unwanted options that
fall under the heading of 'policy' for a specific kernel type.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Use the available ehci and uhci usb features and remove the associated
config settings from the top-level common-pc feature.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Use the available ehci and uhci usb features and remove the associated
config settings from the top-level common-pc feature.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
SMP support was configured in recipe-space, remove it from
recipe-space and move it here.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
SMP support was configured in recipe-space, remove it from
recipe-space and move it here.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
SMP support was configured in recipe-space, remove it from
recipe-space and move it here.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
The boot-live options are already included via common-pc-64 so remove
the associated config settings from the top-level feature.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
The boot-live options are already included via common-pc-64 so remove
the associated config settings from the top-level feature.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Use the available usb features and remove the associated config
settings from the top-level feature.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Use the available usb features and remove the associated config
settings from the top-level feature.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Use the available usb features and remove the associated config
settings from the top-level feature.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Use the available usb features and remove the associated config
settings from the top-level feature.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Use the available usb features and remove the associated config
settings from the top-level crownbay feature.
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
|
|
Move the smp and minimal efi fragments into the base
BSP description rather than the recipe KERNEL_FEATURES.
Drop the -standard features from the base BSP description.
These are already defined in sys940x-standard.scc, and they
may not be wanted in other variants (such as tiny).
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
|
|
Boot fails with a kernel panic in setup_trampoline or shortly
after Freeing kernel memory, when using 0x1000000 for PHYSICAL_START
and PHYSICAL_ALIGN. Use 0x200000 instead.
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
|
|
Add the features commonly added to standard and preempt-rt scc files for
sys940x as well. This enables live boot capability as well various analysis
bits.
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
|
|
Enable efi-ext in the standard and preempt-rt configs so the recipes
do not have to do this explicitly.
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
|
|
Make sys940x consistent with crownbay and fri2 by moving the
emgd-1.10 includes and merges to the $bsp-standard.scc files.
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
|
|
The efi-ext.scc fragment is not part of the standard or preempt-rt
standard features. It is instead a BSP specific feature added to
these ktypes. Rearrange the meta-data accordingly.
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
|
|
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|