| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
The "cmma=" kernel command line parameter needs to be parsed early for
upcoming changes. Therefore move the parsing code.
Note that EX_TABLE handling of cmma_test_essa() needs to be open-coded,
since the early boot code doesn't have infrastructure for handling expected
exceptions.
Reviewed-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
|
|
Use control register bit defines instead of plain numbers where
possible.
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
|
|
Add local and system prefix to some functions to clarify they change
control register contents on either the local CPU or the on all CPUs.
This results in the following API:
Two defines which load and save multiple control registers.
The defines correlate with the following C prototypes:
void __local_ctl_load(unsigned long *, unsigned int cr_low, unsigned int cr_high);
void __local_ctl_store(unsigned long *, unsigned int cr_low, unsigned int cr_high);
Two functions which locally set or clear one bit for a specified
control register:
void local_ctl_set_bit(unsigned int cr, unsigned int bit);
void local_ctl_clear_bit(unsigned int cr, unsigned int bit);
Two functions which set or clear one bit for a specified control
register on all CPUs:
void system_ctl_set_bit(unsigned int cr, unsigned int bit);
void system_ctl_clear_bit(unsigend int cr, unsigned int bit);
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
|
|
The kernel mapping is setup in two stages: in the decompressor map all
pages with RWX permissions, and within the kernel change all mappings to
their final permissions, where most of the mappings are changed from RWX to
RWNX.
Change this and map all pages RWNX from the beginning, however without
enabling noexec via control register modification. This means that
effectively all pages are used with RWX permissions like before. When the
final permissions have been applied to the kernel mapping enable noexec via
control register modification.
This allows to remove quite a bit of non-obvious code.
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
Do the same like x86 with commit 76ea0025a214 ("x86/cpu: Remove "noexec"")
and remove the "noexec" kernel command line option.
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
Make multi-line comment style consistent across the source.
Acked-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
Real Memory Copy and (absolute) Lowcore areas are
not accounted when virtual memory layout is set up.
Fixes: 4df29d2b9024 ("s390/smp: rework absolute lowcore access")
Fixes: 2f0e8aae26a2 ("s390/mm: rework memcpy_real() to avoid DAT-off mode")
Acked-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
Make Real Memory Copy area size and mask explicit.
This does not bring any functional change and only
needed for clarity.
Acked-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
The separate vmalloc area size check against _REGION2_SIZE
is needed in case user provided insanely large value using
vmalloc= kernel command line parameter. That could lead to
overflow and selecting 3 page table levels instead of 4.
Use size_add() for the overflow check and get rid of the
extra vmalloc area check.
With the current values of CONFIG_MAX_PHYSMEM_BITS and
PAGES_PER_SECTION the sum of maximal possible size of
identity mapping and vmemmap area (derived from these
macros) plus modules area size MODULES_LEN can not
overflow. Thus, that sum is used as first addend while
vmalloc area size is second addend for size_add().
Suggested-by: Heiko Carstens <hca@linux.ibm.com>
Acked-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
There are no users of VMEM_MAX_PHYS macro left, remove it.
Reviewed-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
As per description in mm/memory_hotplug.c platforms should define
arch_get_mappable_range() that provides maximum possible addressable
physical memory range for which the linear mapping could be created.
The current implementation uses VMEM_MAX_PHYS macro as the maximum
mappable physical address and it is simply a cast to vmemmap. Since
the address is in physical address space the natural upper limit of
MAX_PHYSMEM_BITS is honoured:
vmemmap_start = min(vmemmap_start, 1UL << MAX_PHYSMEM_BITS);
Further, to make sure the identity mapping would not overlay with
vmemmap, the size of identity mapping could be stripped like this:
ident_map_size = min(ident_map_size, vmemmap_start);
Similarily, any other memory that could be added (e.g DCSS segment)
should not overlay with vmemmap as well and that is prevented by
using vmemmap (VMEM_MAX_PHYS macro) as the upper limit.
However, while the use of VMEM_MAX_PHYS brings the desired result
it actually poses two issues:
1. As described, vmemmap is handled as a physical address, although
it is actually a pointer to struct page in virtual address space.
2. As vmemmap is a virtual address it could have been located
anywhere in the virtual address space. However, the desired
necessity to honour MAX_PHYSMEM_BITS limit prevents that.
Rework arch_get_mappable_range() callback in a way it does not
use VMEM_MAX_PHYS macro and does not confuse the notion of virtual
vs physical address spacees as result. That paves the way for moving
vmemmap elsewhere and optimizing the virtual address space layout.
Introduce max_mappable preserved boot variable and let function
setup_kernel_memory_layout() set it up. As result, the rest of the
code is does not need to know the virtual memory layout specifics.
Reviewed-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
When the KASLR is enabled, randomize the base address of the amode31 image
within the first 2 GB, similar to the approach taken for the vmlinux
image. This makes it harder to predict the location of amode31 data
and code.
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
|
|
Improve the distribution algorithm of random base address to ensure
a uniformity among all suitable addresses. To generate a random value
once, and to build a continuous range in which every value is suitable,
count all the suitable addresses (referred to as positions) that can be
used as a base address. The positions are counted by iterating over the
usable memory ranges. For each range that is big enough to accommodate
the image, count all the suitable addresses where the image can be placed,
while taking reserved memory ranges into consideration.
A new function "iterate_valid_positions()" has dual purpose. Firstly, it
is called to count the positions in a given memory range, and secondly,
to convert a random position back to an address.
"get_random_base()" has been replaced with more generic
"randomize_within_range()" which now could be called for randomizing
base addresses not just for the kernel image.
Acked-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
|
|
The special amode31 part of the kernel must always remain below 2Gb. Place
it just under vmlinux.default_lma by default, which makes it easier to
debug amode31 as its default lma is known 0x10000 - 0x3000 (currently,
amode31's size is 3 pages). This location is always available as it is
originally occupied by the vmlinux archive.
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
|
|
The current modification of the default_lma is illogical and should be
avoided. It would be more appropriate to introduce and utilize a new
variable vmlinux_lma instead, so that default_lma remains unchanged and
at its original "default" value of 0x100000.
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
|
|
Just like other architectures provide a kaslr_enabled() function, instead
of directly accessing a global variable.
Also pass the renamed __kaslr_enabled variable from the decompressor to the
kernel, so that kalsr_enabled() is available there too. This will be used
by a subsequent patch which randomizes the module base load address.
Reviewed-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
|
|
Since regular paging structs are initialized in decompressor already
move KASAN shadow mapping to decompressor as well. This helps to avoid
allocating KASAN required memory in 1 large chunk, de-duplicate paging
structs creation code and start the uncompressed kernel with KASAN
instrumentation right away. This also allows to avoid all pitfalls
accidentally calling KASAN instrumented code during KASAN initialization.
Acked-by: Heiko Carstens <hca@linux.ibm.com>
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
Currently several approaches for finding unused memory in decompressor
are utilized. While "safe_addr" grows towards higher addresses, vmem
code allocates paging structures top down. The former requires careful
ordering. In addition to that ipl report handling code verifies potential
intersections with secure boot certificates on its own. Neither of two
approaches are memory holes aware and consistent with each other in low
memory conditions.
To solve that, existing approaches are generalized and combined
together, as well as online memory ranges are now taken into
consideration.
physmem_info has been extended to contain reserved memory ranges. New
set of functions allow to handle reserves and find unused memory.
All reserves and memory allocations are "typed". In case of out of
memory condition decompressor fails with detailed info on current
reserved ranges and usable online memory.
Linux version 6.2.0 ...
Kernel command line: ... mem=100M
Our of memory allocating 100000 bytes 100000 aligned in range 0:5800000
Reserved memory ranges:
0000000000000000 0000000003e33000 DECOMPRESSOR
0000000003f00000 00000000057648a3 INITRD
00000000063e0000 00000000063e8000 VMEM
00000000063eb000 00000000063f4000 VMEM
00000000063f7800 0000000006400000 VMEM
0000000005800000 0000000006300000 KASAN
Usable online memory ranges (info source: sclp read info [3]):
0000000000000000 0000000006400000
Usable online memory total: 6400000 Reserved: 61b10a3 Free: 24ef5d
Call Trace:
(sp:000000000002bd58 [<0000000000012a70>] physmem_alloc_top_down+0x60/0x14c)
sp:000000000002bdc8 [<0000000000013756>] _pa+0x56/0x6a
sp:000000000002bdf0 [<0000000000013bcc>] pgtable_populate+0x45c/0x65e
sp:000000000002be90 [<00000000000140aa>] setup_vmem+0x2da/0x424
sp:000000000002bec8 [<0000000000011c20>] startup_kernel+0x428/0x8b4
sp:000000000002bf60 [<00000000000100f4>] startup_normal+0xd4/0xd4
physmem_alloc_range allows to find free memory in specified range. It
should be used for one time allocations only like finding position for
amode31 and vmlinux.
physmem_alloc_top_down can be used just like physmem_alloc_range, but
it also allows multiple allocations per type and tries to merge sequential
allocations together. Which is useful for paging structures allocations.
If sequential allocations cannot be merged together they are "chained",
allowing easy per type reserved ranges enumeration and migration to
memblock later. Extra "struct reserved_range" allocated for chaining are
not tracked or reserved but rely on the fact that both
physmem_alloc_range and physmem_alloc_top_down search for free memory
only below current top down allocator position. All reserved ranges
should be transferred to memblock before memblock allocations are
enabled.
The startup code has been reordered to delay any memory allocations until
online memory ranges are detected and occupied memory ranges are marked as
reserved to be excluded from follow-up allocations.
Ipl report certificates are a special case, ipl report certificates list
is checked together with other memory reserves until certificates are
saved elsewhere.
KASAN required memory for shadow memory allocation and mapping is reserved
as 1 large chunk which is later passed to KASAN early initialization code.
Acked-by: Heiko Carstens <hca@linux.ibm.com>
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
In preparation to extending mem_detect with additional information like
reserved ranges rename it to more generic physmem_info. This new naming
also help to avoid confusion by using more exact terms like "physmem
online ranges", etc.
Acked-by: Heiko Carstens <hca@linux.ibm.com>
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
check_image_bootable() has been introduced with commit 627c9b62058e
("s390/boot: block uncompressed vmlinux booting attempts") to make sure
that users don't try to boot uncompressed vmlinux ELF image in qemu. It
used to be possible quite some time ago. That commit prevented confusion
with uncompressed vmlinux image starting to boot and even printing
kernel messages until it crashed. Users might have tried to report the
problem without realizing they are doing something which was not intended.
Since commit f1d3c5323772 ("s390/boot: move sclp early buffer from fixed
address in asm to C") check_image_bootable() doesn't function properly
anymore, as well as booting uncompressed vmlinux image in qemu doesn't
really produce any output and crashes. Moving forward it doesn't make
sense to fix check_image_bootable() anymore, so simply remove it.
Acked-by: Alexander Gordeev <agordeev@linux.ibm.com>
Acked-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
Commit bf64f0517e5d ("s390/mem_detect: handle online memory limit
just once") introduced truncation of mem_detect online ranges
based on identity mapping size. For kdump case however the full
set of online memory ranges has to be feed into memblock_physmem_add
so that crashed system memory could be extracted.
Instead of truncating introduce a "usable limit" which is respected by
mem_detect api. Also add extra online memory ranges iterator which still
provides full set of online memory ranges disregarding the "usable limit".
Fixes: bf64f0517e5d ("s390/mem_detect: handle online memory limit just once")
Reported-by: Alexander Egorenkov <egorenar@linux.ibm.com>
Tested-by: Alexander Egorenkov <egorenar@linux.ibm.com>
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
Fixes: bb1520d581a3 ("s390/mm: start kernel with DAT enabled")
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
Facilities setup has to be done after "facilities" command line option
parsing, it might set extra or remove existing facilities bits for
testing purposes.
Fixes: bb1520d581a3 ("s390/mm: start kernel with DAT enabled")
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
Introduce mem_detect_truncate() to cut any online memory ranges above
established identity mapping size, so that mem_detect users wouldn't
have to do it over and over again.
Suggested-by: Alexander Gordeev <agordeev@linux.ibm.com>
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
Allocation of mem_detect extended area was not considered neither
in commit 9641b8cc733f ("s390/ipl: read IPL report at early boot")
nor in commit b2d24b97b2a9 ("s390/kernel: add support for kernel address
space layout randomization (KASLR)"). As a result mem_detect extended
theoretically may overlap with ipl report or randomized kernel image
position. But as mem_detect code will allocate extended area only
upon exceeding 255 online regions (which should alternate with offline
memory regions) it is not seen in practice.
To make sure mem_detect extended area does not overlap with ipl report
or randomized kernel position extend usage of "safe_addr". Make initrd
handling and mem_detect extended area allocation code move it further
right and make KASLR takes in into consideration as well.
Fixes: 9641b8cc733f ("s390/ipl: read IPL report at early boot")
Fixes: b2d24b97b2a9 ("s390/kernel: add support for kernel address space layout randomization (KASLR)")
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
Commit bb1520d581a3 ("s390/mm: start kernel with DAT enabled")
doesn't consider online memory holes due to potential memory offlining
and erroneously creates pgtables for stand-by memory, which bear RW+X
attribute and trigger a warning:
RANGE SIZE STATE REMOVABLE BLOCK
0x0000000000000000-0x0000000c3fffffff 49G online yes 0-48
0x0000000c40000000-0x0000000c7fffffff 1G offline 49
0x0000000c80000000-0x0000000fffffffff 14G online yes 50-63
0x0000001000000000-0x00000013ffffffff 16G offline 64-79
s390/mm: Found insecure W+X mapping at address 0xc40000000
WARNING: CPU: 14 PID: 1 at arch/s390/mm/dump_pagetables.c:142 note_page+0x2cc/0x2d8
Map only online memory ranges which fit within identity mapping limit.
Fixes: bb1520d581a3 ("s390/mm: start kernel with DAT enabled")
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
Move Real Memory Copy Area allocation to the decompressor.
As result, memcpy_real() and memcpy_real_iter() movers
become usable since the very moment the kernel starts.
Reviewed-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
The setup of the kernel virtual address space is spread
throughout the sources, boot stages and config options
like this:
1. The available physical memory regions are queried
and stored as mem_detect information for later use
in the decompressor.
2. Based on the physical memory availability the virtual
memory layout is established in the decompressor;
3. If CONFIG_KASAN is disabled the kernel paging setup
code populates kernel pgtables and turns DAT mode on.
It uses the information stored at step [1].
4. If CONFIG_KASAN is enabled the kernel early boot
kasan setup populates kernel pgtables and turns DAT
mode on. It uses the information stored at step [1].
The kasan setup creates early_pg_dir directory and
directly overwrites swapper_pg_dir entries to make
shadow memory pages available.
Move the kernel virtual memory setup to the decompressor
and start the kernel with DAT turned on right from the
very first istruction. That completely eliminates the
boot phase when the kernel runs in DAT-off mode, simplies
the overall design and consolidates pgtables setup.
The identity mapping is created in the decompressor, while
kasan shadow mappings are still created by the early boot
kernel code.
Share with decompressor the existing kasan memory allocator.
It decreases the size of a newly requested memory block from
pgalloc_pos and ensures that kernel image is not overwritten.
pgalloc_low and pgalloc_pos pointers are made preserved boot
variables for that.
Use the bootdata infrastructure to setup swapper_pg_dir
and invalid_pg_dir directories used by the kernel later.
The interim early_pg_dir directory established by the
kasan initialization code gets eliminated as result.
As the kernel runs in DAT-on mode only the PSW_KERNEL_BITS
define gets PSW_MASK_DAT bit by default. Additionally, the
setup_lowcore_dat_off() and setup_lowcore_dat_on() routines
get merged, since there is no DAT-off mode stage anymore.
The memory mappings are created with RW+X protection that
allows the early boot code setting up all necessary data
and services for the kernel being booted. Just before the
paging is enabled the memory protection is changed to
RO+X for text, RO+NX for read-only data and RW+NX for
kernel data and the identity mapping.
Reviewed-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
Detect and enable memory facilities which is a
prerequisite for pgtables setup in the decompressor.
Reviewed-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
Nathan Chancellor reported several link errors on s390 with
CONFIG_RELOCATABLE disabled, after binutils commit 906f69cf65da ("IBM
zSystems: Issue error for *DBL relocs on misaligned symbols"). The binutils
commit reveals potential miscompiles that might have happened already
before with linker script defined symbols at odd addresses.
A similar bug was recently fixed in the kernel with commit c9305b6c1f52
("s390: fix nospec table alignments").
See https://github.com/ClangBuiltLinux/linux/issues/1747 for an analysis
from Ulich Weigand.
Therefore always build a relocatable kernel to avoid this problem. There is
hardly any use-case for non-relocatable kernels, so this shouldn't be
controversial.
Link: https://github.com/ClangBuiltLinux/linux/issues/1747
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Reported-by: Nathan Chancellor <nathan@kernel.org>
Tested-by: Nathan Chancellor <nathan@kernel.org>
Link: https://lore.kernel.org/r/20221030182202.2062705-1-hca@linux.ibm.com
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
|
|
Function memcpy_real() is an univeral data mover that does not
require DAT mode to be able reading from a physical address.
Its advantage is an ability to read from any address, even
those for which no kernel virtual mapping exists.
Although memcpy_real() is interrupt-safe, there are no handlers
that make use of this function. The compiler instrumentation
have to be disabled and separate no-DAT stack used to allow
execution of the function once DAT mode is disabled.
Rework memcpy_real() to overcome these shortcomings. As result,
data copying (which is primarily reading out a crashed system
memory by a user process) is executed on a regular stack with
enabled interrupts. Also, use of memcpy_real_buf swap buffer
becomes unnecessary and the swapping is eliminated.
The above is achieved by using a fixed virtual address range
that spans a single page and remaps that page repeatedly when
memcpy_real() is called for a particular physical address.
Reviewed-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
|
|
Temporary unsetting of the prefix page in memcpy_absolute() routine
poses a risk of executing code path with unexpectedly disabled prefix
page. This rework avoids the prefix page uninstalling and disabling
of normal and machine check interrupts when accessing the absolute
zero memory.
Although memcpy_absolute() routine can access the whole memory, it is
only used to update the absolute zero lowcore. This rework therefore
introduces a new mechanism for the absolute zero lowcore access and
scraps memcpy_absolute() routine for good.
Instead, an area is reserved in the virtual memory that is used for
the absolute lowcore access only. That area holds an array of 8KB
virtual mappings - one per CPU. Whenever a CPU is brought online, the
corresponding item is mapped to the real address of the previously
installed prefix page.
The absolute zero lowcore access works like this: a CPU calls the
new primitive get_abs_lowcore() to obtain its 8KB mapping as a
pointer to the struct lowcore. Virtual address references to that
pointer get translated to the real addresses of the prefix page,
which in turn gets swapped with the absolute zero memory addresses
due to prefixing. Once the pointer is not needed it must be released
with put_abs_lowcore() primitive:
struct lowcore *abs_lc;
unsigned long flags;
abs_lc = get_abs_lowcore(&flags);
abs_lc->... = ...;
put_abs_lowcore(abs_lc, flags);
To ensure the described mechanism works large segment- and region-
table entries must be avoided for the 8KB mappings. Failure to do
so results in usage of Region-Frame Absolute Address (RFAA) or
Segment-Frame Absolute Address (SFAA) large page fields. In that
case absolute addresses would be used to address the prefix page
instead of the real ones and the prefixing would get bypassed.
Reviewed-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
|
|
This reverts commit 7d06fed77b7d8fc9f6cc41b4e3f2823d32532ad8.
This introduced vmem_mutex locking from vmem_map_4k_page()
function called from smp_reinit_ipl_cpu() with interrupts
disabled. While it is a pre-SMP early initcall no other CPUs
running in parallel nor other code taking vmem_mutex on this
boot stage - it still needs to be fixed.
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
|
|
Temporary unsetting of the prefix page in memcpy_absolute() routine
poses a risk of executing code path with unexpectedly disabled prefix
page. This rework avoids the prefix page uninstalling and disabling
of normal and machine check interrupts when accessing the absolute
zero memory.
Although memcpy_absolute() routine can access the whole memory, it is
only used to update the absolute zero lowcore. This rework therefore
introduces a new mechanism for the absolute zero lowcore access and
scraps memcpy_absolute() routine for good.
Instead, an area is reserved in the virtual memory that is used for
the absolute lowcore access only. That area holds an array of 8KB
virtual mappings - one per CPU. Whenever a CPU is brought online, the
corresponding item is mapped to the real address of the previously
installed prefix page.
The absolute zero lowcore access works like this: a CPU calls the
new primitive get_abs_lowcore() to obtain its 8KB mapping as a
pointer to the struct lowcore. Virtual address references to that
pointer get translated to the real addresses of the prefix page,
which in turn gets swapped with the absolute zero memory addresses
due to prefixing. Once the pointer is not needed it must be released
with put_abs_lowcore() primitive:
struct lowcore *abs_lc;
unsigned long flags;
abs_lc = get_abs_lowcore(&flags);
abs_lc->... = ...;
put_abs_lowcore(abs_lc, flags);
To ensure the described mechanism works large segment- and region-
table entries must be avoided for the 8KB mappings. Failure to do
so results in usage of Region-Frame Absolute Address (RFAA) or
Segment-Frame Absolute Address (SFAA) large page fields. In that
case absolute addresses would be used to address the prefix page
instead of the real ones and the prefixing would get bypassed.
Reviewed-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
|
|
Uncouple input and output arguments by making the latter
the function return value.
Reviewed-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
|
|
The final kernel image is created by linking decompressor object files with
a startup archive. The startup archive file however does not contain only
optional code and data which can be discarded if not referenced. It also
contains mandatory object data like head.o which must never be discarded,
even if not referenced.
Move the decompresser code and linker script to the boot directory and get
rid of the startup archive so everything is kept during link time.
Acked-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
Initial KASAN shadow memory range was picked to preserve original kernel
modules area position. With protected execution support, which might
impose addressing limitation on vmalloc area and hence affect modules
area position, current fixed KASAN shadow memory range is only making
kernel memory layout setup more complex. So move it to the very end of
available virtual space and simplify calculations.
At the same time return to previous kernel address space split. In
particular commit 0c4f2623b957 ("s390: setup kernel memory layout
early") introduced precise identity map size calculation and keeping
vmemmap left most starting from a fresh region table entry. This didn't
take into account additional mapping region requirement for potential
DCSS mapping above available physical memory. So go back to virtual
space split between 1:1 mapping & vmemmap array once vmalloc area size
is subtracted.
Cc: stable@vger.kernel.org
Fixes: 0c4f2623b957 ("s390: setup kernel memory layout early")
Reported-by: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
Reviewed-by: Heiko Carstens <hca@linux.ibm.com>
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
The memory for amode31 section is allocated from the decompressed
kernel. Instead, allocate that memory from the decompressor. This
is a prerequisite to allow initialization of the virtual memory
before the decompressed kernel takes over.
Reviewed-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
|
|
Move offsetting all of vmlinux_info fields to a separate
function for better readability.
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
sparse warning:
CHECK arch/s390/boot/startup.c
arch/s390/boot/startup.c:283:39: error: arithmetics on pointers to functions
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
This change simplifies the task of making the decompressor relocatable.
The decompressor's image contains special DMA sections between _sdma and
_edma. This DMA segment is loaded at boot as part of the decompressor and
then simply handed over to the decompressed kernel. The decompressor itself
never uses it in any way. The primary reason for this is the need to keep
the aforementioned DMA segment below 2GB which is required by architecture,
and because the decompressor is always loaded at a fixed low physical
address, it is guaranteed that the DMA region will not cross the 2GB
memory limit. If the DMA region had been placed in the decompressed kernel,
then KASLR would make this guarantee impossible to fulfill or it would
be restricted to the first 2GB of memory address space.
This commit moves all DMA sections between _sdma and _edma from
the decompressor's image to the decompressed kernel's image. The complete
DMA region is placed in the init section of the decompressed kernel and
immediately relocated below 2GB at start-up before it is needed by other
parts of the decompressed kernel. The relocation of the DMA region happens
even if the decompressed kernel is already located below 2GB in order
to keep the first implementation simple. The relocation should not have
any noticeable impact on boot time because the DMA segment is only a couple
of pages.
After relocating the DMA sections, the kernel has to fix all references
which point into it. In order to automate this, place all variables
pointing into the DMA sections in a special .dma.refs section. All such
variables must be defined using the new __dma_ref macro. Only variables
containing addresses within the DMA sections must be placed in the new
.dma.refs section.
Furthermore, move the initialization of control registers from
the decompressor to the decompressed kernel because some control registers
reference tables that must be placed in the DMA data section to
guarantee that their addresses are below 2G. Because the decompressed
kernel relocates the DMA sections at startup, the content of control
registers CR2, CR5 and CR15 must be updated with new addresses after
the relocation. The decompressed kernel initializes all control registers
early at boot and then updates the content of CR2, CR5 and CR15
as soon as the DMA relocation has occurred. This practically reverts
the commit a80313ff91ab ("s390/kernel: introduce .dma sections").
Signed-off-by: Alexander Egorenkov <egorenar@linux.ibm.com>
Acked-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
The new boot data struct shall replace global variables OLDMEM_BASE and
OLDMEM_SIZE. It is initialized in the decompressor and passed
to the decompressed kernel. In comparison to the old solution, this one
doesn't access data at fixed physical addresses which will become important
when the decompressor becomes relocatable.
Signed-off-by: Alexander Egorenkov <egorenar@linux.ibm.com>
Acked-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
The new boot data struct shall replace global variables INITRD_START and
INITRD_SIZE. It is initialized in the decompressor and passed
to the decompressed kernel. In comparison to the old solution, this one
doesn't access data at fixed physical addresses which will become important
when the decompressor becomes relocatable.
Signed-off-by: Alexander Egorenkov <egorenar@linux.ibm.com>
Acked-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
A dump kernel is neither required nor able to support Secure Execution.
Signed-off-by: Alexander Egorenkov <egorenar@linux.ibm.com>
Reviewed-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
The functions adjust_to_uv_max() and uv_query_info() are used only
in the decompressor. Therefore, move the function declarations from
the global arch/s390/include/asm/uv.h to arch/s390/boot/uv.h.
Signed-off-by: Alexander Egorenkov <egorenar@linux.ibm.com>
Reviewed-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
To prevent multiple incompatible declarations of symbols and to catch
such mistakes at compile time.
Signed-off-by: Alexander Egorenkov <egorenar@linux.ibm.com>
Acked-by: Heiko Carstens <hca@linux.ibm.com>
Acked-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
|
The magic string "S390EP" at offset 0x10008 indicated to the decompressed
kernel that it was booted by the decompressor. Introduce a new bootdata
flag instead which conveys the same information in an explicit and
a cleaner way. But keep the magic string because it is a kernel ABI.
Signed-off-by: Alexander Egorenkov <egorenar@linux.ibm.com>
Reviewed-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
|
|
Currently there are two separate places where kernel memory layout has
to be known and adjusted:
1. early kasan setup.
2. paging setup later.
Those 2 places had to be kept in sync and adjusted to reflect peculiar
technical details of one another. With additional factors which influence
kernel memory layout like ultravisor secure storage limit, complexity
of keeping two things in sync grew up even more.
Besides that if we look forward towards creating identity mapping and
enabling DAT before jumping into uncompressed kernel - that would also
require full knowledge of and control over kernel memory layout.
So, de-duplicate and move kernel memory layout setup logic into
the decompressor.
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
|
|
With gcc-11, there are a lot of warnings because the facility functions
are accessing lowcore through a null pointer. Fix this by moving the
facility arrays away from lowcore.
Signed-off-by: Sven Schnelle <svens@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
|
|
Currently we have to consider too many different values which
in the end only affect identity mapping size. These are:
1. max_physmem_end - end of physical memory online or standby.
Always <= end of the last online memory block (get_mem_detect_end()).
2. CONFIG_MAX_PHYSMEM_BITS - the maximum size of physical memory the
kernel is able to support.
3. "mem=" kernel command line option which limits physical memory usage.
4. OLDMEM_BASE which is a kdump memory limit when the kernel is executed as
crash kernel.
5. "hsa" size which is a memory limit when the kernel is executed during
zfcp/nvme dump.
Through out kernel startup and run we juggle all those values at once
but that does not bring any amusement, only confusion and complexity.
Unify all those values to a single one we should really care, that is
our identity mapping size.
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Acked-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
|
