ARM: config: Refresh IXP4xx config after multiplatform
After converting to multiplatform the old IXP4xx defconfig
doesn't even activate the ethernet driver anymore, we need
to refresh this thoroughly.
- Drop options that ARCH_MULTI_V5 selects for us.
- Some options moved around because of multiplatform and
because generic Kconfig movements.
- Drop all MACH_* that do not exist anymore.
- Compress the kernel and initramfs with XZ as the memory
is often limited for these machines.
- Make sure IXP4XX_QMGR and IXP4XX_NPE is selected.
- Make sure the MTD_PHYSMAP_IXP4XX is selected and not
the old MTD_IXP4XX (will be deleted).
- Activate CRYPTO_DEV_IXP4XX
- Add some very basic default drivers such as AT24 EEPROMs
used in many routers.
- Add some kernel configs needed to bring up OpenWrt
which is the only userspace for these devices:
CGROUPS, DEVTMPFS, INOTIFY_USER, SIGNALFD, TIMERFD,
EPOLL, OVERLAY_FS and SQUASHFS.
One of my multiplatform patches went a little too far and removed
a declaration that is needed for compile-testing the omapfb
driver on non-OMAP1 platforms:
arm-linux-gnueabi-ld: drivers/video/fbdev/omap/omapfb_main.o: in function `omapfb_do_probe':
omapfb_main.c:(.text+0x41ec): undefined reference to `omap_set_dma_priority'
Add back the inline stub, and in turn hide the definition when
omapfb is disabled, like we do for the usb specific bits.
Reported-by: kernel test robot <lkp@intel.com> Reviewed-by: Tony Lindgren <tony@atomide.com> Fixes: c6b7d9d359f4 ("dma: omap: hide legacy interface") Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Moving the plat-omap files triggered a sparse warning for omap1 and
omap2 that is now in a different file from before. Found some more sparse
warnings here that I address by making sure the declaration is visible
to both the caller and the callee, or they are static
mach-omap1/fb.c:33:17: warning: symbol 'omap_fb_resources' was not declared. Should it be static?
mach-omap1/timer32k.c:215:12: warning: symbol 'omap_init_clocksource_32k' was not declared. Should it be static?
mach-omap1/i2c.c:36:12: warning: symbol 'omap_i2c_add_bus' was not declared. Should it be static?
mach-omap1/i2c.c:115:12: warning: symbol 'omap_register_i2c_bus_cmdline' was not declared. Should it be static?
mach-omap1/i2c.c:140:12: warning: symbol 'omap_register_i2c_bus' was not declared. Should it be static?
mach-omap2/dma.c:180:34: warning: symbol 'dma_plat_info' was not declared. Should it be static?
mach-omap2/omap4-common.c:116:6: warning: symbol 'omap_interconnect_sync' was not declared. Should it be static?
mach-omap2/omap-iommu.c:113:5: warning: symbol 'omap_iommu_set_pwrdm_constraint' was not declared. Should it be static?
Reported-by: kernel test robot <lkp@intel.com> Reviewed-by: Tony Lindgren <tony@atomide.com> Tested-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Two of the remaining four headers are never used, the other
two are only used inside of the platform code. Move them
next to the files that include them and remove the
Makefile trick that adds the include path.
Merge branch 'omap1/multiplatform-prep' of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc into arm/multiplatform
This is the full series for converting OMAP1 to multiplatform, rebased
from my 2019 attempt to do the same thing. The soc tree contains simpler
patches to do the same for iop32x, ixp4xx, ep93xx and s3c24xx, which
means we are getting closer to completing this for all ARMv5 platforms
(I have patches for PXA, which is the last one remaining).
Janusz already tested the branch separately and did the missing work
for the common-clk conversion after my previous approach was broken.
Aaro found one regression during additional testing, this is fixed now.
Acked-by: Tony Lindgren <tony@atomide.com> Tested-by: Aaro Koskinen <aaro.koskinen@iki.fi> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Tony Lindgren [Wed, 13 Apr 2022 08:16:56 +0000 (11:16 +0300)]
MAINTAINERS: omap1: Add Janusz as an additional maintainer
Janusz has been active with improving and testing the omap1 SoC support
and has been recently working on adding support for the common clock
framework.
Signed-off-by: Tony Lindgren <tony@atomide.com> Acked-by: Janusz Krzysztofik <jmkrzyszt@gmail.com> Acked-by: Aaro Koskinen <aaro.koskinen@iki.fi> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
There are some OMAP1 clock code bits that have no effect:
- crystal_type variable is set to 0 but never changed, then
crystal_type == 2 condition is never true and ck_ref.rate never set to 19200000,
- clk->ops->allow_idle() is called from omap_clk_enable_autoidle_all() but
that op is not configured for any clock, then the function does nothing
and the op field is not needed,
- ENABLE_ON_INIT flag is set for some clocks but is never checked by any
code, then not needed.
Drop that code.
Signed-off-by: Janusz Krzysztofik <jmkrzyszt@gmail.com> Acked-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
The code of OMAP1 clocks contains quite a few unused elements:
- functions and function like macros never called: clk_reparent(),
recalculate_root_clocks(), clk_enable_init_clocks(),
omap_clk_get_by_name(), omap_clk_disable_autoidle_all(),
__clk_get_parent(clk), __clk_get_rate(),
- unused structure fields:
- clkops: .find_idlest(), .find_companion(), .deny_idle(),
- clk: .src_offset, as well as .clkdm -- no longer present but still
mentioned in comments,
- definitions of unused flags: INVERT_ENABLE, CLOCK_CLKOUTX2,
- definitions of unused data types: struct clk_functions,
- prototypes of functions with no implementation: clk_init(),
omap1_watchdog_recalc().
- declarations of never defined global variables: clkops_dummy.
Drop them.
Signed-off-by: Janusz Krzysztofik <jmkrzyszt@gmail.com> Acked-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Since its introduction to the mainline kernel, omap1_uart_recalc() helper
makes incorrect use of clk->enable_bit as a ready to use bitmap mask while
it only provides the bit number. Fix it.
Signed-off-by: Janusz Krzysztofik <jmkrzyszt@gmail.com> Acked-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Commit d6149910b282 ("ARM: OMAP: Fix CONFIG_DEBUG_LL") was supposed to fix
low level debugging, most possibly by early enabling UART clocks. The fix
actually introduced early reset of most bits of MOD_CONF_CTRL_0 register,
with the exception of UART1 and UART2 clock related bits which were set
high. However, UART1 clock bit can play different roles on different
OMAP1 variants. On OMAP1610 it enables the clock as intended, but on
OMAP1510 it switches the clock rate from 12 to 48 MHz. Even worth, for
UART2 the bit changes its clock rate also on OMAP1610. As a result, UART
rates set by a bootloader can be unintentionally changed early on kernel
boot and low level debugging broken, not fixed. Besides, reset of all
other bits was not justified.
Don't touch register bits not related to UART clocks. Also, don't touch
the bit of UART2 clock. Make sure UART1 and UART3 are enabled early on
relevant OMAP1610 machine types while preserving bootloader UART clock
rates on others.
Signed-off-by: Janusz Krzysztofik <jmkrzyszt@gmail.com> Acked-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
ARM: OMAP1: Prepare for conversion of OMAP1 clocks to CCF
In preparation for conversion of OMAP1 clocks to common clock framework,
identify arch/arm/mach-omap1 local users of those clocks and update them
to call clk_prepare_enable/clk_disable_unprepare() instead of just
clk_enable/disable(), as required by CCF implementation of clock API.
Signed-off-by: Janusz Krzysztofik <jmkrzyszt@gmail.com> Acked-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
In a multiplatform randconfig kernel, one can have
CONFIG_ARCH_OMAP1 enabled, but none of the specific SoCs.
This leads to some build issues as the code is not
meant to deal with this configuration at the moment:
arch/arm/mach-omap1/io.c:86:20: error: unused function 'omap1_map_common_io' [-Werror,-Wunused-function]
arch/arm/mach-omap1/pm.h:113:2: error: "Power management for this processor not implemented yet" [-Werror,-W#warnings]
Use the same trick as on OMAP2 and guard the actual compilation
of platform code with another Makefile ifdef check based
on an option that depends on having at least one SoC enabled.
The io.c file still needs to get compiled to allow building
device drivers with a dependency on CONFIG_ARCH_OMAP1.
Acked-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Arnd Bergmann [Tue, 6 Aug 2019 14:16:03 +0000 (16:16 +0200)]
ARM: omap1: move mach/*.h into mach directory
Most of the header files are no longer referenced from outside
arch/arm/mach-omap1, so move them all to that place directly
and change their users to use the new location.
The exceptions are:
- mach/compress.h is used by the core architecture code
- mach/serial.h is used by mach/compress.h
The mach/memory.h is empty and gets removed in the process,
avoiding the need for CONFIG_NEED_MACH_MEMORY_H.
Acked-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
The sram initialization code is the only shared omap1/2 code that
is not a standalone driver, but it is very short. Having two copies
of this code means some duplication of the sources, but actually
saves object code size as it can be inlined better.
Acked-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
The legacy interface for omap-dma is only used on OMAP1, and the
same is true for the non-DT case. Make both of these conditional on
CONFIG_ARCH_OMAP1 being set to simplify the dependency.
The non-OMAP stub functions in include/linux/omap-dma.h are note needed
any more either now, because they are only called on OMAP1.
The Samsung S3C24xx and S3C64xx platforms are very old designs. S3C2416
was introduced in 2008 and S3C6410 in 2009/2010. They are not widely
available anymore - out-of-stock on FriendlyArm (one of manufacturers of
boards) and only few specialist stores still offer them for quite a high
price.
The community around these platforms was not very active, so I suspect
no one really uses them anymore. Maintenance takes precious time so
there is little sense in keeping them alive if there are no real users.
Let's mark all S3C24xx and S3C64xx platforms as deprecated and mention
possible removal in after 2022 for the first and 2024 for the lattere.
The deprecation message will be as text in Kconfig, build message (not a
warning though) and runtime print error.
If there are any users, they might respond and postpone the removal.
Arnd Bergmann [Fri, 27 Nov 2020 14:46:47 +0000 (15:46 +0100)]
ARM: omap: dma: make usb support optional
Most of the plat-omap/dma.c code is specific to the USB
driver. Hide that code when it is not in use, to make it
clearer which parts are actually still required.
Acked-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
ARM: omap1: move 32k counter from plat-omap to mach-omap1
omap2 stopped using this code with commit d8ff95327a0c ("ARM: OMAP2+:
Remove unused legacy code for timer"), so just move it to mach-omap1 now,
along with the other half of that driver.
Acked-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Arnd Bergmann [Tue, 6 Aug 2019 14:55:58 +0000 (16:55 +0200)]
input: omap: void using mach/*.h headers
The omap-keypad driver currently relies on including mach/memory.h
implicitly, but that won't happen once omap1 is converted to
CONFIG_ARCH_MULTIPLATFORM. Include the required header
explicitly.
Arnd Bergmann [Tue, 6 Aug 2019 14:18:29 +0000 (16:18 +0200)]
clocksource: ti-dmtimer: avoid using mach/hardware.h
As a preparation for future omap1 multiplatform support, stop
using mach/hardware.h and instead include the omap1-io.h
for low-level register access to MOD_CONF_CTRL_1.
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org> Acked-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Arnd Bergmann [Tue, 6 Aug 2019 14:10:57 +0000 (16:10 +0200)]
usb: omap: avoid mach/*.h headers
The omap usb drivers still rely on mach/*.h headers that
are explicitly or implicitly included, but all the required
definitions are now in include/linux/soc/ti/, so use those
instead and allow compile-testing on other architectures.
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Acked-by: Felipe Balbi <felipe.balbi@linux.intel.com> Acked-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Arnd Bergmann [Mon, 5 Aug 2019 21:08:58 +0000 (23:08 +0200)]
ARM: omap1: move CF chipselect setup to board file
There is only one board that uses the omap_cf driver, so
moving the chipselect configuration there does not lead
to code duplication but avoids the use of mach/tc.h
in drivers.
Acked-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Arnd Bergmann [Tue, 6 Aug 2019 13:24:58 +0000 (15:24 +0200)]
ARM: omap1: move some headers to include/linux/soc
There are three remaining header files that are used by omap1
specific device drivers:
- mach/soc.h provides cpu_is_omapXXX abstractions
- mach/hardware.h provides omap_read/omap_write functions
and physical addresses
- mach/mux.h provides an omap specific pinctrl abstraction
This is generally not how we do platform abstractions today,
and it would be good to completely get rid of these in favor
of passing information through platform devices and the pinctrl
subsystem.
However, given that nobody is working on that, just move it
one step forward by splitting out the header files that are
used by drivers today from the machine headers that are only
used internally.
Acked-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Arnd Bergmann [Tue, 6 Aug 2019 12:34:31 +0000 (14:34 +0200)]
ARM: omap1: move mach/usb.h to include/linux/soc
The register definitions in this header are used in at least four
different places, with little hope of completely cleaning that up.
Split up the file into a portion that becomes a linux-wide header
under include/linux/soc/ti/, and the parts that are actually only
needed by board files.
Acked-by: Felipe Balbi <felipe.balbi@linux.intel.com> Acked-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Arnd Bergmann [Tue, 24 Nov 2020 20:19:05 +0000 (21:19 +0100)]
ARM: omap1: ams-delta: remove camera leftovers
The obsolete camera support was removed, but a few lines remain in this
file and cause a warning:
arch/arm/mach-omap1/board-ams-delta.c:462:12: warning: 'ams_delta_camera_power' defined but not used [-Wunused-function]
462 | static int ams_delta_camera_power(struct device *dev, int power)
Arnd Bergmann [Fri, 9 Aug 2019 19:27:01 +0000 (21:27 +0200)]
fbdev: omap: pass irqs as resource
To avoid relying on the mach/irqs.h header, stop using
OMAP_LCDC_IRQ and INT_1610_SoSSI_MATCH directly in the driver
code, but instead pass these as resources.
Arnd Bergmann [Fri, 9 Aug 2019 19:42:31 +0000 (21:42 +0200)]
ARM: omap1: declare a dummy omap_set_dma_priority
omapfb calls directly into the omap_set_dma_priority() function in
the DMA driver. This prevents compile-testing omapfb on other
architectures. Add an inline function next to the other ones
for non-omap configurations.
Arnd Bergmann [Mon, 5 Aug 2019 21:15:37 +0000 (23:15 +0200)]
ARM: omap1: move lcd_dma code into omapfb driver
The omapfb driver is split into platform specific code for omap1, and
driver code that is also specific to omap1.
Moving both parts into the driver directory simplifies the structure
and avoids the dependency on certain omap machine header files.
As mach/lcd_dma.h can not be included from include/linux/omap-dma.h
any more now, move the omap_lcd_dma_running() declaration into the
omap-dma header, which matches where it is defined.
A recent cleanup patch removed the only reference to a local variable
in some configurations.
Move the variable into the one block it is still used in, inside
of an #ifdef, to avoid this warning.
Fixes: 1eedaa46be26 ("video: fbdev: omapfb: lcd_ams_delta: Make use of the helper function dev_err_probe()") Acked-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
After all the work that Linus Walleij did on this platform, it can be
part of a generic kernel build as well.
Note that there are known bugs in little-endian mode on ixp4xx, and
no other ARMv5 platform at this point supports big-endian mode, or is
likely to in the future, so there is limited practical value in this,
but it helps with build testing and ixp4xx little-endian support may
get fixed in the future.
Choosing big-endian vs little-endian kernels in Kconfig has not worked
correctly since the introduction of CONFIG_ARCH_MULTIPLATFORM a long
time ago.
The problems is that CONFIG_BIG_ENDIAN depends on
ARCH_SUPPORTS_BIG_ENDIAN, which can set by any one platform
in the config, but would actually have to be supported by all
of them.
This was mostly ok for ARMv6/ARMv7 builds, since these are BE8 and
tend to just work aside from problems in nonportable device drivers.
For ARMv4/v5 machines, CONFIG_BIG_ENDIAN and CONFIG_ARCH_MULTIPLATFORM
were never set together, so this was disabled on all those machines
except for IXP4xx.
As IXP4xx can now become part of ARCH_MULTIPLATFORM, it seems better to
formalize this logic: all ARMv4/v5 platforms get an explicit dependency
on being either big-endian (ixp4xx) or little-endian (the rest). We may
want to fix ixp4xx in the future to support both, but it does not work
in LE mode at the moment.
For the ARMv6/v7 platforms, there are two ways this could be handled
a) allow both modes only for platforms selecting
'ARCH_SUPPORTS_BIG_ENDIAN' today, but only LE mode for the
others, given that these were added intentionally at some
point.
b) allow both modes everwhere, given that it was already possible
to build that way by e.g. selecting ARCH_VIRT, and that the
list is not an accurate reflection of which platforms may or
may not work.
Out of these, I picked b) because it seemed slighly more logical
to me.
After iop32x was converted to the generic multi-irq entry
code, nothing really stops us from building it into a
generic kernel.
The two last headers can simply be removed, the mach/irqs.h
gets replaced with the sparse-irq intiialization from the
board specific .nr_irqs value, and the decompressor debug
output can use the debug_ll hack that all other platforms
use.
With the custom ISA I/O and the missing sparse-irq support
out of the way, s3c24xx can now be built into the same
kernel as all other ARM9 based platforms.
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
As a final bit of preparation for converting to ARCH_MULTIPLATFORM,
change the interrupt handling for s3c24xx to use sparse IRQs.
Since the number of possible interrupts is already fixed and relatively
small per chip, just make it use all legacy interrupts preallocated
using the .nr_irqs field in the machine descriptor, rather than actually
allocating domains on the fly.
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
ARM: s3c24xx: remove support for ISA drivers on BAST PC/104
BAST is the one machine that theoretically supports unmodified ISA
drivers for hardware on its PC/104 connector, using a custom version of
the inb()/outb() and inw()/outw() macros.
This is incompatible with the generic version used in asm/io.h, and
can't easily be used in a multiplatform kernel.
Removing the special case for 16-bit I/O port access on BAST gets us
closer to multiplatform, at the expense of any PC/104 users with 16-bit
cards having to either use an older kernel or modify their ISA drivers
to manually ioremap() the area and use readw()/write() in place of
inw()/outw(). Either way is probably ok, given that there is a
recurring discussion about dropping s3c24xx altogether, and many
traditional ISA drivers are already gone.
Machines other than BAST already have no support for ISA drivers, though a
couple of them do map one of the external chip-selects into the ISA port
range, using the same address for 8-bit and 16-bit I/O. It is unlikely
that anything actually uses this mapping, but it's also easy to keep
this working by mapping it to the normal platform-independent PCI I/O
base that is otherwise unused on s3c24xx.
The mach/map-base.h file is no longer referenced in global headers and
can be moved into the platform directory.
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Arnd Bergmann [Sun, 23 Feb 2014 10:30:13 +0000 (11:30 +0100)]
ARM: dove: multiplatform support
The dove platform is now ready to be enabled for multiplatform
support, this patch does the switch over by modifying the Kconfig file,
the defconfig and removing the last mach/*.h header that becomes obsolete
with this.
This work was originally done in 2015 as all the ARMv7 machiens
gove moved over to multiplatform builds, but at the time it conflicted
with some patches that Russell was trying to upstream, so we
left it at that.
I hope that there is no longer a need to keep dove separate from the
rest, so we can either add it to the other ARMv7 platforms, or just
replace it with the DT based platform code for the same hardware
in mach-mvebu and remove mach-dove entirely.
Acked-by: Andrew Lunn <andrew@lunn.ch> Cc: Russell King <linux@armlinux.org.uk> Cc: Jason Cooper <jason@lakedaemon.net> Cc: Sebastian Hesselbarth <sebastian.hesselbarth@gmail.com> Cc: Gregory Clement <gregory.clement@bootlin.com> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
With the clock support and the interrupts out of the way, ep93xx can be
compiled into the same kernel image as the other ARMv4/v5 platforms. The
last obstacle are the two workarounds for broken boot loaders that
require us to re-initialize the ethernet controller and/or the watchdog
on certain machines.
Move this code into the decompressor sources directly, checking for
each possibly affected machine individually.
Without CONFIG_SPARSE_IRQ, we rely on mach/irqs.h to define NR_IRQS
globally. Do the minimal conversion by setting .nr_irqs in each
machine descriptor.
Only the vision_ep9307 machine has extra IRQs for GPIOs, so make
.nr_irqs the original value there, while using the plain NR_EP93XX_IRQS
everywhere else.
Merge tag 'trace-v5.18-2' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace
Pull more tracing updates from Steven Rostedt:
- Rename the staging files to give them some meaning. Just
stage1,stag2,etc, does not show what they are for
- Check for NULL from allocation in bootconfig
- Hold event mutex for dyn_event call in user events
- Mark user events to broken (to work on the API)
- Remove eBPF updates from user events
- Remove user events from uapi header to keep it from being installed.
- Move ftrace_graph_is_dead() into inline as it is called from hot
paths and also convert it into a static branch.
* tag 'trace-v5.18-2' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:
tracing: Move user_events.h temporarily out of include/uapi
ftrace: Make ftrace_graph_is_dead() a static branch
tracing: Set user_events to BROKEN
tracing/user_events: Remove eBPF interfaces
tracing/user_events: Hold event_mutex during dyn_event_add
proc: bootconfig: Add null pointer check
tracing: Rename the staging files for trace_events
Merge tag 'clk-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux
Pull clk fix from Stephen Boyd:
"A single revert to fix a boot regression seen when clk_put() started
dropping rate range requests. It's best to keep various systems
booting so we'll kick this out and try again next time"
* tag 'clk-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux:
Revert "clk: Drop the rate range on clk_put()"
Merge tag 'x86-urgent-2022-04-03' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull x86 fixes from Thomas Gleixner:
"A set of x86 fixes and updates:
- Make the prctl() for enabling dynamic XSTATE components correct so
it adds the newly requested feature to the permission bitmap
instead of overwriting it. Add a selftest which validates that.
- Unroll string MMIO for encrypted SEV guests as the hypervisor
cannot emulate it.
- Handle supervisor states correctly in the FPU/XSTATE code so it
takes the feature set of the fpstate buffer into account. The
feature sets can differ between host and guest buffers. Guest
buffers do not contain supervisor states. So far this was not an
issue, but with enabling PASID it needs to be handled in the buffer
offset calculation and in the permission bitmaps.
- Avoid a gazillion of repeated CPUID invocations in by caching the
values early in the FPU/XSTATE code.
- Enable CONFIG_WERROR in x86 defconfig.
- Make the X86 defconfigs more useful by adapting them to Y2022
reality"
* tag 'x86-urgent-2022-04-03' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/fpu/xstate: Consolidate size calculations
x86/fpu/xstate: Handle supervisor states in XSTATE permissions
x86/fpu/xsave: Handle compacted offsets correctly with supervisor states
x86/fpu: Cache xfeature flags from CPUID
x86/fpu/xsave: Initialize offset/size cache early
x86/fpu: Remove unused supervisor only offsets
x86/fpu: Remove redundant XCOMP_BV initialization
x86/sev: Unroll string mmio with CC_ATTR_GUEST_UNROLL_STRING_IO
x86/config: Make the x86 defconfigs a bit more usable
x86/defconfig: Enable WERROR
selftests/x86/amx: Update the ARCH_REQ_XCOMP_PERM test
x86/fpu/xstate: Fix the ARCH_REQ_XCOMP_PERM implementation
Merge tag 'core-urgent-2022-04-03' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull RT signal fix from Thomas Gleixner:
"Revert the RT related signal changes. They need to be reworked and
generalized"
* tag 'core-urgent-2022-04-03' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
Revert "signal, x86: Delay calling signals in atomic on RT enabled kernels"
Merge tag 'dma-mapping-5.18-1' of git://git.infradead.org/users/hch/dma-mapping
Pull more dma-mapping updates from Christoph Hellwig:
- fix a regression in dma remap handling vs AMD memory encryption (me)
- finally kill off the legacy PCI DMA API (Christophe JAILLET)
* tag 'dma-mapping-5.18-1' of git://git.infradead.org/users/hch/dma-mapping:
dma-mapping: move pgprot_decrypted out of dma_pgprot
PCI/doc: cleanup references to the legacy PCI DMA API
PCI: Remove the deprecated "pci-dma-compat.h" API
Stephen Boyd [Sun, 3 Apr 2022 02:28:18 +0000 (19:28 -0700)]
Revert "clk: Drop the rate range on clk_put()"
This reverts commit 0bc6a8728d95e710c7f6cb09aa25958dfa6f63db. There are
multiple reports that this breaks boot on various systems. The common
theme is that orphan clks are having rates set on them when that isn't
expected. Let's revert it out for now so that -rc1 boots.
Merge tag 'perf-tools-for-v5.18-2022-04-02' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux
Pull more perf tools updates from Arnaldo Carvalho de Melo:
- Avoid SEGV if core.cpus isn't set in 'perf stat'.
- Stop depending on .git files for building PERF-VERSION-FILE, used in
'perf --version', fixing some perf tools build scenarios.
- Convert tracepoint.py example to python3.
- Update UAPI header copies from the kernel sources: socket,
mman-common, msr-index, KVM, i915 and cpufeatures.
- Update copy of libbpf's hashmap.c.
- Directly return instead of using local ret variable in
evlist__create_syswide_maps(), found by coccinelle.
* tag 'perf-tools-for-v5.18-2022-04-02' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux:
perf python: Convert tracepoint.py example to python3
perf evlist: Directly return instead of using local ret variable
perf cpumap: More cpu map reuse by merge.
perf cpumap: Add is_subset function
perf evlist: Rename cpus to user_requested_cpus
perf tools: Stop depending on .git files for building PERF-VERSION-FILE
tools headers cpufeatures: Sync with the kernel sources
tools headers UAPI: Sync drm/i915_drm.h with the kernel sources
tools headers UAPI: Sync linux/kvm.h with the kernel sources
tools kvm headers arm64: Update KVM headers from the kernel sources
tools arch x86: Sync the msr-index.h copy with the kernel sources
tools headers UAPI: Sync asm-generic/mman-common.h with the kernel
perf beauty: Update copy of linux/socket.h with the kernel sources
perf tools: Update copy of libbpf's hashmap.c
perf stat: Avoid SEGV if core.cpus isn't set
Merge tag 'kbuild-fixes-v5.18' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild
Pull Kbuild fixes from Masahiro Yamada:
- Fix empty $(PYTHON) expansion.
- Fix UML, which got broken by the attempt to suppress Clang warnings.
- Fix warning message in modpost.
* tag 'kbuild-fixes-v5.18' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild:
modpost: restore the warning message for missing symbol versions
Revert "um: clang: Strip out -mno-global-merge from USER_CFLAGS"
kbuild: Remove '-mno-global-merge'
kbuild: fix empty ${PYTHON} in scripts/link-vmlinux.sh
kconfig: remove stale comment about removed kconfig_print_symbol()
Merge tag 'mips_5.18_1' of git://git.kernel.org/pub/scm/linux/kernel/git/mips/linux
Pull MIPS fixes from Thomas Bogendoerfer:
- build fix for gpio
- fix crc32 build problems
- check for failed memory allocations
* tag 'mips_5.18_1' of git://git.kernel.org/pub/scm/linux/kernel/git/mips/linux:
MIPS: crypto: Fix CRC32 code
MIPS: rb532: move GPIOD definition into C-files
MIPS: lantiq: check the return value of kzalloc()
mips: sgi-ip22: add a check for the return of kzalloc()
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
Pull kvm fixes from Paolo Bonzini:
- Only do MSR filtering for MSRs accessed by rdmsr/wrmsr
- Documentation improvements
- Prevent module exit until all VMs are freed
- PMU Virtualization fixes
- Fix for kvm_irq_delivery_to_apic_fast() NULL-pointer dereferences
- Other miscellaneous bugfixes
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (42 commits)
KVM: x86: fix sending PV IPI
KVM: x86/mmu: do compare-and-exchange of gPTE via the user address
KVM: x86: Remove redundant vm_entry_controls_clearbit() call
KVM: x86: cleanup enter_rmode()
KVM: x86: SVM: fix tsc scaling when the host doesn't support it
kvm: x86: SVM: remove unused defines
KVM: x86: SVM: move tsc ratio definitions to svm.h
KVM: x86: SVM: fix avic spec based definitions again
KVM: MIPS: remove reference to trap&emulate virtualization
KVM: x86: document limitations of MSR filtering
KVM: x86: Only do MSR filtering when access MSR by rdmsr/wrmsr
KVM: x86/emulator: Emulate RDPID only if it is enabled in guest
KVM: x86/pmu: Fix and isolate TSX-specific performance event logic
KVM: x86: mmu: trace kvm_mmu_set_spte after the new SPTE was set
KVM: x86/svm: Clear reserved bits written to PerfEvtSeln MSRs
KVM: x86: Trace all APICv inhibit changes and capture overall status
KVM: x86: Add wrappers for setting/clearing APICv inhibits
KVM: x86: Make APICv inhibit reasons an enum and cleanup naming
KVM: X86: Handle implicit supervisor access with SMAP
KVM: X86: Rename variable smap to not_smap in permission_fault()
...
Merge tag 'for-5.18/drivers-2022-04-02' of git://git.kernel.dk/linux-block
Pull block driver fix from Jens Axboe:
"Got two reports on nbd spewing warnings on load now, which is a
regression from a commit that went into your tree yesterday.
Revert the problematic change for now"
* tag 'for-5.18/drivers-2022-04-02' of git://git.kernel.dk/linux-block:
Revert "nbd: fix possible overflow on 'first_minor' in nbd_dev_add()"
Merge tag 'tag-chrome-platform-for-v5.18' of git://git.kernel.org/pub/scm/linux/kernel/git/chrome-platform/linux
Pull chrome platform updates from Benson Leung:
"cros_ec_typec:
- Check for EC device - Fix a crash when using the cros_ec_typec
driver on older hardware not capable of typec commands
- Make try power role optional
- Mux configuration reorganization series from Prashant
cros_ec_debugfs:
- Fix use after free. Thanks Tzung-bi
sensorhub:
- cros_ec_sensorhub fixup - Split trace include file
misc:
- Add new mailing list for chrome-platform development:
chrome-platform@lists.linux.dev
Now with patchwork!"
* tag 'tag-chrome-platform-for-v5.18' of git://git.kernel.org/pub/scm/linux/kernel/git/chrome-platform/linux:
platform/chrome: cros_ec_debugfs: detach log reader wq from devm
platform: chrome: Split trace include file
platform/chrome: cros_ec_typec: Update mux flags during partner removal
platform/chrome: cros_ec_typec: Configure muxes at start of port update
platform/chrome: cros_ec_typec: Get mux state inside configure_mux
platform/chrome: cros_ec_typec: Move mux flag checks
platform/chrome: cros_ec_typec: Check for EC device
platform/chrome: cros_ec_typec: Make try power role optional
MAINTAINERS: platform-chrome: Add new chrome-platform@lists.linux.dev list
After being merged, user_events become more visible to a wider audience
that have concerns with the current API.
It is too late to fix this for this release, but instead of a full
revert, just mark it as BROKEN (which prevents it from being selected in
make config). Then we can work finding a better API. If that fails,
then it will need to be completely reverted.
To not have the code silently bitrot, still allow building it with
COMPILE_TEST.
And to prevent the uapi header from being installed, then later changed,
and then have an old distro user space see the old version, move the
header file out of the uapi directory.
Surround the include with CONFIG_COMPILE_TEST to the current location,
but when the BROKEN tag is taken off, it will use the uapi directory,
and fail to compile. This is a good way to remind us to move the header
back.
tracing: Move user_events.h temporarily out of include/uapi
While user_events API is under development and has been marked for broken
to not let the API become fixed, move the header file out of the uapi
directory. This is to prevent it from being installed, then later changed,
and then have an old distro user space update with a new kernel, where
applications see the user_events being available, but the old header is in
place, and then they get compiled incorrectly.
Also, surround the include with CONFIG_COMPILE_TEST to the current
location, but when the BROKEN tag is taken off, it will use the uapi
directory, and fail to compile. This is a good way to remind us to move
the header back.
Christophe Leroy [Wed, 30 Mar 2022 07:00:19 +0000 (09:00 +0200)]
ftrace: Make ftrace_graph_is_dead() a static branch
ftrace_graph_is_dead() is used on hot paths, it just reads a variable
in memory and is not worth suffering function call constraints.
For instance, at entry of prepare_ftrace_return(), inlining it avoids
saving prepare_ftrace_return() parameters to stack and restoring them
after calling ftrace_graph_is_dead().
While at it using a static branch is even more performant and is
rather well adapted considering that the returned value will almost
never change.
Inline ftrace_graph_is_dead() and replace 'kill_ftrace_graph' bool
by a static branch.
After being merged, user_events become more visible to a wider audience
that have concerns with the current API. It is too late to fix this for
this release, but instead of a full revert, just mark it as BROKEN (which
prevents it from being selected in make config). Then we can work finding
a better API. If that fails, then it will need to be completely reverted.
tracing: Rename the staging files for trace_events
When looking for implementation of different phases of the creation of the
TRACE_EVENT() macro, it is pretty useless when all helper macro
redefinitions are in files labeled "stageX_defines.h". Rename them to
state which phase the files are for. For instance, when looking for the
defines that are used to create the event fields, seeing
"stage4_event_fields.h" gives the developer a good idea that the defines
are in that file.
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Li RongQing [Wed, 9 Mar 2022 08:35:44 +0000 (16:35 +0800)]
KVM: x86: fix sending PV IPI
If apic_id is less than min, and (max - apic_id) is greater than
KVM_IPI_CLUSTER_SIZE, then the third check condition is satisfied but
the new apic_id does not fit the bitmask. In this case __send_ipi_mask
should send the IPI.
This is mostly theoretical, but it can happen if the apic_ids on three
iterations of the loop are for example 1, KVM_IPI_CLUSTER_SIZE, 0.
Fixes: 3c7773d7663 ("KVM: X86: Implement PV IPIs in linux guest") Signed-off-by: Li RongQing <lirongqing@baidu.com>
Message-Id: <1646814944-51801-1-git-send-email-lirongqing@baidu.com> Cc: stable@vger.kernel.org Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Paolo Bonzini [Tue, 29 Mar 2022 16:56:24 +0000 (12:56 -0400)]
KVM: x86/mmu: do compare-and-exchange of gPTE via the user address
FNAME(cmpxchg_gpte) is an inefficient mess. It is at least decent if it
can go through get_user_pages_fast(), but if it cannot then it tries to
use memremap(); that is not just terribly slow, it is also wrong because
it assumes that the VM_PFNMAP VMA is contiguous.
The right way to do it would be to do the same thing as
hva_to_pfn_remapped() does since commit a59d68b54670 ("KVM: MMU: try to
fix up page faults before giving up", 2016-07-05), using follow_pte()
and fixup_user_fault() to determine the correct address to use for
memremap(). To do this, one could for example extract hva_to_pfn()
for use outside virt/kvm/kvm_main.c. But really there is no reason to
do that either, because there is already a perfectly valid address to
do the cmpxchg() on, only it is a userspace address. That means doing
user_access_begin()/user_access_end() and writing the code in assembly
to handle exceptions correctly. Worse, the guest PTE can be 8-byte
even on i686 so there is the extra complication of using cmpxchg8b to
account for. But at least it is an efficient mess.
(Thanks to Linus for suggesting improvement on the inline assembly).
Reported-by: Qiuhao Li <qiuhao@sysec.org> Reported-by: Gaoning Pan <pgn@zju.edu.cn> Reported-by: Yongkang Jia <kangel@zju.edu.cn> Reported-by: syzbot+6cde2282daa792c49ab8@syzkaller.appspotmail.com Debugged-by: Tadeusz Struk <tadeusz.struk@linaro.org> Tested-by: Maxim Levitsky <mlevitsk@redhat.com> Cc: stable@vger.kernel.org Fixes: b89f9ae1559d ("X86/KVM: Handle PFNs outside of kernel reach when touching GPTEs") Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
When emulating exit from long mode, EFER_LMA is cleared with
vmx_set_efer(). This will already unset the VM_ENTRY_IA32E_MODE control
bit as requested by SDM, so there is no need to unset VM_ENTRY_IA32E_MODE
again in exit_lmode() explicitly. In case EFER isn't supported by
hardware, long mode isn't supported, so exit_lmode() cannot be reached.
Note that, thanks to the shadow controls mechanism, this change doesn't
eliminate vmread or vmwrite.
Signed-off-by: Zhenzhong Duan <zhenzhong.duan@intel.com>
Message-Id: <20220311102643.807507-3-zhenzhong.duan@intel.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Zhenzhong Duan [Fri, 11 Mar 2022 10:26:42 +0000 (18:26 +0800)]
KVM: x86: cleanup enter_rmode()
vmx_set_efer() sets uret->data but, in fact if the value of uret->data
will be used vmx_setup_uret_msrs() will have rewritten it with the value
returned by update_transition_efer(). uret->data is consumed if and only
if uret->load_into_hardware is true, and vmx_setup_uret_msrs() takes care
of (a) updating uret->data before setting uret->load_into_hardware to true
(b) setting uret->load_into_hardware to false if uret->data isn't updated.
Opportunistically use "vmx" directly instead of redoing to_vmx().
Signed-off-by: Zhenzhong Duan <zhenzhong.duan@intel.com>
Message-Id: <20220311102643.807507-2-zhenzhong.duan@intel.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Paolo Bonzini [Sun, 13 Mar 2022 14:05:22 +0000 (15:05 +0100)]
KVM: MIPS: remove reference to trap&emulate virtualization
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220313140522.1307751-1-pbonzini@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Paolo Bonzini [Tue, 15 Mar 2022 22:17:15 +0000 (18:17 -0400)]
KVM: x86: document limitations of MSR filtering
MSR filtering requires an exit to userspace that is hard to implement and
would be very slow in the case of nested VMX vmexit and vmentry MSR
accesses. Document the limitation.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Hou Wenlong [Mon, 7 Mar 2022 12:26:33 +0000 (20:26 +0800)]
KVM: x86: Only do MSR filtering when access MSR by rdmsr/wrmsr
If MSR access is rejected by MSR filtering,
kvm_set_msr()/kvm_get_msr() would return KVM_MSR_RET_FILTERED,
and the return value is only handled well for rdmsr/wrmsr.
However, some instruction emulation and state transition also
use kvm_set_msr()/kvm_get_msr() to do msr access but may trigger
some unexpected results if MSR access is rejected, E.g. RDPID
emulation would inject a #UD but RDPID wouldn't cause a exit
when RDPID is supported in hardware and ENABLE_RDTSCP is set.
And it would also cause failure when load MSR at nested entry/exit.
Since msr filtering is based on MSR bitmap, it is better to only
do MSR filtering for rdmsr/wrmsr.
Signed-off-by: Hou Wenlong <houwenlong.hwl@antgroup.com>
Message-Id: <2b2774154f7532c96a6f04d71c82a8bec7d9e80b.1646655860.git.houwenlong.hwl@antgroup.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Hou Wenlong [Wed, 2 Mar 2022 13:15:14 +0000 (21:15 +0800)]
KVM: x86/emulator: Emulate RDPID only if it is enabled in guest
When RDTSCP is supported but RDPID is not supported in host,
RDPID emulation is available. However, __kvm_get_msr() would
only fail when RDTSCP/RDPID both are disabled in guest, so
the emulator wouldn't inject a #UD when RDPID is disabled but
RDTSCP is enabled in guest.
Like Xu [Wed, 9 Mar 2022 08:42:57 +0000 (16:42 +0800)]
KVM: x86/pmu: Fix and isolate TSX-specific performance event logic
HSW_IN_TX* bits are used in generic code which are not supported on
AMD. Worse, these bits overlap with AMD EventSelect[11:8] and hence
using HSW_IN_TX* bits unconditionally in generic code is resulting in
unintentional pmu behavior on AMD. For example, if EventSelect[11:8]
is 0x2, pmc_reprogram_counter() wrongly assumes that
HSW_IN_TX_CHECKPOINTED is set and thus forces sampling period to be 0.
Also per the SDM, both bits 32 and 33 "may only be set if the processor
supports HLE or RTM" and for "IN_TXCP (bit 33): this bit may only be set
for IA32_PERFEVTSEL2."
Opportunistically eliminate code redundancy, because if the HSW_IN_TX*
bit is set in pmc->eventsel, it is already set in attr.config.
Reported-by: Ravi Bangoria <ravi.bangoria@amd.com> Reported-by: Jim Mattson <jmattson@google.com> Fixes: bbe6638bf4f6 ("perf, kvm: Support the in_tx/in_tx_cp modifiers in KVM arch perfmon emulation v5") Co-developed-by: Ravi Bangoria <ravi.bangoria@amd.com> Signed-off-by: Ravi Bangoria <ravi.bangoria@amd.com> Signed-off-by: Like Xu <likexu@tencent.com>
Message-Id: <20220309084257.88931-1-likexu@tencent.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
projects / kernel.git / log