linux-stable.git/arch/arm/kernel/setup.c, branch v4.14.331 Linux kernel stable tree ARM: 9081/1: fix gcc-10 thumb2-kernel regression 2021-06-30T12:48:56+00:00 Arnd Bergmann arnd@arndb.de 2021-05-14T10:26:37+00:00 af115fa3d6fc530f10f3045d889ee4488c756675 commit dad7b9896a5dbac5da8275d5a6147c65c81fb5f2 upstream. When building the kernel wtih gcc-10 or higher using the CONFIG_CC_OPTIMIZE_FOR_PERFORMANCE=y flag, the compiler picks a slightly different set of registers for the inline assembly in cpu_init() that subsequently results in a corrupt kernel stack as well as remaining in FIQ mode. If a banked register is used for the last argument, the wrong version of that register gets loaded into CPSR_c. When building in Arm mode, the arguments are passed as immediate values and the bug cannot happen. This got introduced when Daniel reworked the FIQ handling and was technically always broken, but happened to work with both clang and gcc before gcc-10 as long as they picked one of the lower registers. This is probably an indication that still very few people build the kernel in Thumb2 mode. Marek pointed out the problem on IRC, Arnd narrowed it down to this inline assembly and Russell pinpointed the exact bug. Change the constraints to force the final mode switch to use a non-banked register for the argument to ensure that the correct constant gets loaded. Another alternative would be to always use registers for the constant arguments to avoid the #ifdef that has now become more complex. Cc: <stable@vger.kernel.org> # v3.18+ Cc: Daniel Thompson <daniel.thompson@linaro.org> Reported-by: Marek Vasut <marek.vasut@gmail.com> Acked-by: Ard Biesheuvel <ardb@kernel.org> Fixes: c0e7f7ee717e ("ARM: 8150/3: fiq: Replace default FIQ handler") Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
commit dad7b9896a5dbac5da8275d5a6147c65c81fb5f2 upstream.

When building the kernel wtih gcc-10 or higher using the
CONFIG_CC_OPTIMIZE_FOR_PERFORMANCE=y flag, the compiler picks a slightly
different set of registers for the inline assembly in cpu_init() that
subsequently results in a corrupt kernel stack as well as remaining in
FIQ mode. If a banked register is used for the last argument, the wrong
version of that register gets loaded into CPSR_c.  When building in Arm
mode, the arguments are passed as immediate values and the bug cannot
happen.

This got introduced when Daniel reworked the FIQ handling and was
technically always broken, but happened to work with both clang and gcc
before gcc-10 as long as they picked one of the lower registers.
This is probably an indication that still very few people build the
kernel in Thumb2 mode.

Marek pointed out the problem on IRC, Arnd narrowed it down to this
inline assembly and Russell pinpointed the exact bug.

Change the constraints to force the final mode switch to use a non-banked
register for the argument to ensure that the correct constant gets loaded.
Another alternative would be to always use registers for the constant
arguments to avoid the #ifdef that has now become more complex.

Cc: <stable@vger.kernel.org> # v3.18+
Cc: Daniel Thompson <daniel.thompson@linaro.org>
Reported-by: Marek Vasut <marek.vasut@gmail.com>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Fixes: c0e7f7ee717e ("ARM: 8150/3: fiq: Replace default FIQ handler")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
ARM: spectre-v2: per-CPU vtables to work around big.Little systems 2019-02-20T09:20:47+00:00 Russell King rmk+kernel@armlinux.org.uk 2019-02-14T02:10:20+00:00 89229bc38d9c98e0cde2c83a7e7e360ed1400911 Commit 383fb3ee8024d596f488d2dbaf45e572897acbdb upstream. In big.Little systems, some CPUs require the Spectre workarounds in paths such as the context switch, but other CPUs do not. In order to handle these differences, we need per-CPU vtables. We are unable to use the kernel's per-CPU variables to support this as per-CPU is not initialised at times when we need access to the vtables, so we have to use an array indexed by logical CPU number. We use an array-of-pointers to avoid having function pointers in the kernel's read/write .data section. Reviewed-by: Julien Thierry <julien.thierry@arm.com> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk> Signed-off-by: David A. Long <dave.long@linaro.org> Reviewed-by: Julien Thierry <julien.thierry@arm.com> Tested-by: Julien Thierry <julien.thierry@arm.com> Signed-off-by: Sasha Levin <sashal@kernel.org> 
Commit 383fb3ee8024d596f488d2dbaf45e572897acbdb upstream.

In big.Little systems, some CPUs require the Spectre workarounds in
paths such as the context switch, but other CPUs do not.  In order
to handle these differences, we need per-CPU vtables.

We are unable to use the kernel's per-CPU variables to support this
as per-CPU is not initialised at times when we need access to the
vtables, so we have to use an array indexed by logical CPU number.

We use an array-of-pointers to avoid having function pointers in
the kernel's read/write .data section.

Reviewed-by: Julien Thierry <julien.thierry@arm.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David A. Long <dave.long@linaro.org>
Reviewed-by: Julien Thierry <julien.thierry@arm.com>
Tested-by: Julien Thierry <julien.thierry@arm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ARM: add PROC_VTABLE and PROC_TABLE macros 2019-02-20T09:20:47+00:00 Russell King rmk+kernel@armlinux.org.uk 2019-02-14T02:10:19+00:00 94944687acf7d398099c3841a99553c65201df98 Commit e209950fdd065d2cc46e6338e47e52841b830cba upstream. Allow the way we access members of the processor vtable to be changed at compile time. We will need to move to per-CPU vtables to fix the Spectre variant 2 issues on big.Little systems. However, we have a couple of calls that do not need the vtable treatment, and indeed cause a kernel warning due to the (later) use of smp_processor_id(), so also introduce the PROC_TABLE macro for these which always use CPU 0's function pointers. Reviewed-by: Julien Thierry <julien.thierry@arm.com> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk> Signed-off-by: David A. Long <dave.long@linaro.org> Reviewed-by: Julien Thierry <julien.thierry@arm.com> Tested-by: Julien Thierry <julien.thierry@arm.com> Signed-off-by: Sasha Levin <sashal@kernel.org> 
Commit e209950fdd065d2cc46e6338e47e52841b830cba upstream.

Allow the way we access members of the processor vtable to be changed
at compile time.  We will need to move to per-CPU vtables to fix the
Spectre variant 2 issues on big.Little systems.

However, we have a couple of calls that do not need the vtable
treatment, and indeed cause a kernel warning due to the (later) use
of smp_processor_id(), so also introduce the PROC_TABLE macro for
these which always use CPU 0's function pointers.

Reviewed-by: Julien Thierry <julien.thierry@arm.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David A. Long <dave.long@linaro.org>
Reviewed-by: Julien Thierry <julien.thierry@arm.com>
Tested-by: Julien Thierry <julien.thierry@arm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ARM: split out processor lookup 2019-02-20T09:20:46+00:00 Russell King rmk+kernel@armlinux.org.uk 2019-02-14T02:10:17+00:00 48b874ec26b9e8c92cd0284754fe22b96c4deddd Commit 65987a8553061515b5851b472081aedb9837a391 upstream. Split out the lookup of the processor type and associated error handling from the rest of setup_processor() - we will need to use this in the secondary CPU bringup path for big.Little Spectre variant 2 mitigation. Reviewed-by: Julien Thierry <julien.thierry@arm.com> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk> Signed-off-by: David A. Long <dave.long@linaro.org> Reviewed-by: Julien Thierry <julien.thierry@arm.com> Tested-by: Julien Thierry <julien.thierry@arm.com> Signed-off-by: Sasha Levin <sashal@kernel.org> 
Commit 65987a8553061515b5851b472081aedb9837a391 upstream.

Split out the lookup of the processor type and associated error handling
from the rest of setup_processor() - we will need to use this in the
secondary CPU bringup path for big.Little Spectre variant 2 mitigation.

Reviewed-by: Julien Thierry <julien.thierry@arm.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David A. Long <dave.long@linaro.org>
Reviewed-by: Julien Thierry <julien.thierry@arm.com>
Tested-by: Julien Thierry <julien.thierry@arm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ARM: kexec: avoid allocating crashkernel region outside lowmem 2017-07-20T10:37:15+00:00 Russell King rmk+kernel@armlinux.org.uk 2017-07-19T22:01:38+00:00 67556d7a851c20116923c23f1d49ecdec954e3a0 Allocating the crashkernel region outside lowmem causes the kernel to oops while trying to kexec into the new kernel: Loading crashdump kernel... Unable to handle kernel NULL pointer dereference at virtual address 00000000 pgd = edd70000 [00000000] *pgd=de19e835 Internal error: Oops: 817 [#2] SMP ARM Modules linked in: ... CPU: 0 PID: 689 Comm: sh Not tainted 4.12.0-rc3-next-20170601-04015-gc3a5a20 Hardware name: Generic DRA74X (Flattened Device Tree) task: edb32f00 task.stack: edf18000 PC is at memcpy+0x50/0x330 LR is at 0xe3c34001 pc : [<c04baf30>] lr : [<e3c34001>] psr: 800c0193 sp : edf19c2c ip : 0a000001 fp : c0553170 r10: c055316e r9 : 00000001 r8 : e3130001 r7 : e4903004 r6 : 0a000014 r5 : e3500000 r4 : e59f106c r3 : e59f0074 r2 : ffffffe8 r1 : c010fb88 r0 : 00000000 Flags: Nzcv IRQs off FIQs on Mode SVC_32 ISA ARM Segment none Control: 10c5387d Table: add7006a DAC: 00000051 Process sh (pid: 689, stack limit = 0xedf18218) Stack: (0xedf19c2c to 0xedf1a000) ... [<c04baf30>] (memcpy) from [<c010fae0>] (machine_kexec+0xa8/0x12c) [<c010fae0>] (machine_kexec) from [<c01e4104>] (__crash_kexec+0x5c/0x98) [<c01e4104>] (__crash_kexec) from [<c01e419c>] (crash_kexec+0x5c/0x68) [<c01e419c>] (crash_kexec) from [<c010c5c0>] (die+0x228/0x490) [<c010c5c0>] (die) from [<c011e520>] (__do_kernel_fault.part.0+0x54/0x1e4) [<c011e520>] (__do_kernel_fault.part.0) from [<c082412c>] (do_page_fault+0x1e8/0x400) [<c082412c>] (do_page_fault) from [<c010135c>] (do_DataAbort+0x38/0xb8) [<c010135c>] (do_DataAbort) from [<c0823584>] (__dabt_svc+0x64/0xa0) This is caused by image->control_code_page being a highmem page, so page_address(image->control_code_page) returns NULL. In any case, we don't want the control page to be a highmem page. We already limit the crash kernel region to the top of 32-bit physical memory space. Also limit it to the top of lowmem in physical space. Reported-by: Keerthy <j-keerthy@ti.com> Tested-by: Keerthy <j-keerthy@ti.com> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk> 
Allocating the crashkernel region outside lowmem causes the kernel to
oops while trying to kexec into the new kernel:

Loading crashdump kernel...
Unable to handle kernel NULL pointer dereference at virtual address 00000000
pgd = edd70000
[00000000] *pgd=de19e835
Internal error: Oops: 817 [#2] SMP ARM
Modules linked in: ...
CPU: 0 PID: 689 Comm: sh Not tainted 4.12.0-rc3-next-20170601-04015-gc3a5a20
Hardware name: Generic DRA74X (Flattened Device Tree)
task: edb32f00 task.stack: edf18000
PC is at memcpy+0x50/0x330
LR is at 0xe3c34001
pc : [<c04baf30>]    lr : [<e3c34001>]    psr: 800c0193
sp : edf19c2c  ip : 0a000001  fp : c0553170
r10: c055316e  r9 : 00000001  r8 : e3130001
r7 : e4903004  r6 : 0a000014  r5 : e3500000  r4 : e59f106c
r3 : e59f0074  r2 : ffffffe8  r1 : c010fb88  r0 : 00000000
Flags: Nzcv  IRQs off  FIQs on  Mode SVC_32  ISA ARM  Segment none
Control: 10c5387d  Table: add7006a  DAC: 00000051
Process sh (pid: 689, stack limit = 0xedf18218)
Stack: (0xedf19c2c to 0xedf1a000)
...
[<c04baf30>] (memcpy) from [<c010fae0>] (machine_kexec+0xa8/0x12c)
[<c010fae0>] (machine_kexec) from [<c01e4104>] (__crash_kexec+0x5c/0x98)
[<c01e4104>] (__crash_kexec) from [<c01e419c>] (crash_kexec+0x5c/0x68)
[<c01e419c>] (crash_kexec) from [<c010c5c0>] (die+0x228/0x490)
[<c010c5c0>] (die) from [<c011e520>] (__do_kernel_fault.part.0+0x54/0x1e4)
[<c011e520>] (__do_kernel_fault.part.0) from [<c082412c>] (do_page_fault+0x1e8/0x400)
[<c082412c>] (do_page_fault) from [<c010135c>] (do_DataAbort+0x38/0xb8)
[<c010135c>] (do_DataAbort) from [<c0823584>] (__dabt_svc+0x64/0xa0)

This is caused by image->control_code_page being a highmem page, so
page_address(image->control_code_page) returns NULL.  In any case, we
don't want the control page to be a highmem page.

We already limit the crash kernel region to the top of 32-bit physical
memory space.  Also limit it to the top of lowmem in physical space.

Reported-by: Keerthy <j-keerthy@ti.com>
Tested-by: Keerthy <j-keerthy@ti.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
ARM: 8682/1: V7M: Set cacheid iff DminLine or IminLine is nonzero 2017-06-12T14:47:29+00:00 Vladimir Murzin vladimir.murzin@arm.com 2017-06-12T12:35:52+00:00 d360a687d99577110c181e67ebfb9a1b6fed63a2 Cache support is optional feature in M-class cores, thus DminLine or IminLine of Cache Type Register is zero if caches are not implemented, but we check the whole CTR which has other features encoded there. Let's be more precise and check for DminLine and IminLine of CTR before we set cacheid. Signed-off-by: Vladimir Murzin <vladimir.murzin@arm.com> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk> 
Cache support is optional feature in M-class cores, thus DminLine or
IminLine of Cache Type Register is zero if caches are not implemented,
but we check the whole CTR which has other features encoded there.
Let's be more precise and check for DminLine and IminLine of CTR
before we set cacheid.

Signed-off-by: Vladimir Murzin <vladimir.murzin@arm.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
ARM: 8667/3: Fix memory attribute inconsistencies when using fixmap 2017-04-20T10:19:42+00:00 Jon Medhurst tixy@linaro.org 2017-04-10T10:13:59+00:00 b089c31c519c3906c14801b6ec483e18a5152a50 To cope with the variety in ARM architectures and configurations, the pagetable attributes for kernel memory are generated at runtime to match the system the kernel finds itself on. This calculated value is stored in pgprot_kernel. However, when early fixmap support was added for ARM (commit a5f4c561b3b1) the attributes used for mappings were hard coded because pgprot_kernel is not set up early enough. Unfortunately, when fixmap is used after early boot this means the memory being mapped can have different attributes to existing mappings, potentially leading to unpredictable behaviour. A specific problem also exists due to the hard coded values not include the 'shareable' attribute which means on systems where this matters (e.g. those with multiple CPU clusters) the cache contents for a memory location can become inconsistent between CPUs. To resolve these issues we change fixmap to use the same memory attributes (from pgprot_kernel) that the rest of the kernel uses. To enable this we need to refactor the initialisation code so build_mem_type_table() is called early enough. Note, that relies on early param parsing for memory type overrides passed via the kernel command line, so we need to make sure this call is still after parse_early_params(). [ardb: keep early_fixmap_init() before param parsing, for earlycon] Fixes: a5f4c561b3b1 ("ARM: 8415/1: early fixmap support for earlycon") Cc: <stable@vger.kernel.org> # v4.3+ Tested-by: afzal mohammed <afzal.mohd.ma@gmail.com> Signed-off-by: Jon Medhurst <tixy@linaro.org> Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk> 
To cope with the variety in ARM architectures and configurations, the
pagetable attributes for kernel memory are generated at runtime to match
the system the kernel finds itself on. This calculated value is stored
in pgprot_kernel.

However, when early fixmap support was added for ARM (commit
a5f4c561b3b1) the attributes used for mappings were hard coded because
pgprot_kernel is not set up early enough. Unfortunately, when fixmap is
used after early boot this means the memory being mapped can have
different attributes to existing mappings, potentially leading to
unpredictable behaviour. A specific problem also exists due to the hard
coded values not include the 'shareable' attribute which means on
systems where this matters (e.g. those with multiple CPU clusters) the
cache contents for a memory location can become inconsistent between
CPUs.

To resolve these issues we change fixmap to use the same memory
attributes (from pgprot_kernel) that the rest of the kernel uses. To
enable this we need to refactor the initialisation code so
build_mem_type_table() is called early enough. Note, that relies on early
param parsing for memory type overrides passed via the kernel command
line, so we need to make sure this call is still after
parse_early_params().

[ardb: keep early_fixmap_init() before param parsing, for earlycon]

Fixes: a5f4c561b3b1 ("ARM: 8415/1: early fixmap support for earlycon")
Cc: <stable@vger.kernel.org> # v4.3+
Tested-by: afzal mohammed <afzal.mohd.ma@gmail.com>
Signed-off-by: Jon Medhurst <tixy@linaro.org>
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
ARM: 8637/1: Adjust memory boundaries after reservations 2017-02-28T11:05:28+00:00 Laura Abbott labbott@redhat.com 2017-01-13T21:51:45+00:00 985626564eedc470ce2866e53938303368ad41b7 adjust_lowmem_bounds is responsible for setting up the boundary for lowmem/highmem. This needs to be setup before memblock reservations can occur. At the time memblock reservations can occur, memory can also be removed from the system. The lowmem/highmem boundary and end of memory may be affected by this but it is currently not recalculated. On some systems this may be harmless, on others this may result in incorrect ranges being passed to the main memory allocator. Correct this by recalculating the lowmem/highmem boundary after all reservations have been made. Tested-by: Magnus Lilja <lilja.magnus@gmail.com> Signed-off-by: Laura Abbott <labbott@redhat.com> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk> 
adjust_lowmem_bounds is responsible for setting up the boundary for
lowmem/highmem. This needs to be setup before memblock reservations can
occur. At the time memblock reservations can occur, memory can also be
removed from the system. The lowmem/highmem boundary and end of memory
may be affected by this but it is currently not recalculated. On some
systems this may be harmless, on others this may result in incorrect
ranges being passed to the main memory allocator. Correct this by
recalculating the lowmem/highmem boundary after all reservations have
been made.

Tested-by: Magnus Lilja <lilja.magnus@gmail.com>
Signed-off-by: Laura Abbott <labbott@redhat.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
ARM: 8636/1: Cleanup sanity_check_meminfo 2017-02-28T11:04:57+00:00 Laura Abbott labbott@redhat.com 2017-01-13T21:51:08+00:00 374d446d25d6271ee615952a3b7f123ba4983c35 The logic for sanity_check_meminfo has become difficult to follow. Clean up the code so it's more obvious what the code is actually trying to do. Additionally, meminfo is now removed so rename the function to better describe its purpose. Tested-by: Magnus Lilja <lilja.magnus@gmail.com> Reviewed-by: Nicolas Pitre <nico@linaro.org> Signed-off-by: Laura Abbott <lauraa@codeaurora.org> Signed-off-by: Laura Abbott <labbott@redhat.com> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk> 
The logic for sanity_check_meminfo has become difficult to
follow. Clean up the code so it's more obvious what the code
is actually trying to do. Additionally, meminfo is now removed
so rename the function to better describe its purpose.

Tested-by: Magnus Lilja <lilja.magnus@gmail.com>
Reviewed-by: Nicolas Pitre <nico@linaro.org>
Signed-off-by: Laura Abbott <lauraa@codeaurora.org>
Signed-off-by: Laura Abbott <labbott@redhat.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
ARM: 8604/1: V7M: Add support for reading the CTR with read_cpuid_cachetype() 2016-09-06T14:51:07+00:00 Jonathan Austin jonathan.austin@arm.com 2016-08-30T16:27:19+00:00 f5a5c89e36d0897b65e4e6bc2f646f75f8074263 With the addition of caches to the V7M Architecture a new Cache Type Register (CTR) is defined at 0xE000ED7C. This register serves the same purpose as the V7A/R version and accessed via the read_cpuid_cachetype. Signed-off-by: Jonathan Austin <jonathan.austin@arm.com> Signed-off-by: Vladimir Murzin <vladimir.murzin@arm.com> Tested-by: Andras Szemzo <sza@esh.hu> Tested-by: Joachim Eastwood <manabian@gmail.com> Tested-by: Alexandre TORGUE <alexandre.torgue@st.com> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk> 
With the addition of caches to the V7M Architecture a new Cache Type
Register (CTR) is defined at 0xE000ED7C. This register serves the same
purpose as the V7A/R version and accessed via the read_cpuid_cachetype.

Signed-off-by: Jonathan Austin <jonathan.austin@arm.com>
Signed-off-by: Vladimir Murzin <vladimir.murzin@arm.com>
Tested-by: Andras Szemzo <sza@esh.hu>
Tested-by: Joachim Eastwood <manabian@gmail.com>
Tested-by: Alexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>