Monitoring tools that want to find out which resctrl control and monitor
groups a task belongs to must currently read the "tasks" file in every
group until they locate the process ID.
Add an additional file /proc/{pid}/cpu_resctrl_groups to provide this
information:
1) res:
mon:
resctrl is not available.
2) res:/
mon:
Task is part of the root resctrl control group, and it is not associated
to any monitor group.
3) res:/
mon:mon0
Task is part of the root resctrl control group and monitor group mon0.
4) res:group0
mon:
Task is part of resctrl control group group0, and it is not associated
to any monitor group.
5) res:group0
mon:mon1
Task is part of resctrl control group group0 and monitor group mon1.
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Tested-by: Jinshi Chen <jinshi.chen@intel.com>
Link: https://lkml.kernel.org/r/20200115092851.14761-1-yu.c.chen@intel.com
When checking whether the reported lfb_size makes sense, the height
* stride result is page-aligned before seeing whether it exceeds the
reported size.
This doesn't work if height * stride is not an exact number of pages.
For example, as reported in the kernel bugzilla below, an 800x600x32 EFI
framebuffer gets skipped because of this.
Move the PAGE_ALIGN to after the check vs size.
Reported-by: Christopher Head <chead@chead.ca>
Tested-by: Christopher Head <chead@chead.ca>
Signed-off-by: Arvind Sankar <nivedita@alum.mit.edu>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://bugzilla.kernel.org/show_bug.cgi?id=206051
Link: https://lkml.kernel.org/r/20200107230410.2291947-1-nivedita@alum.mit.edu
The final build stage of the x86 kernel captures some symbol
addresses from the decompressor binary and copies them into zoffset.h.
It uses sed with a regular expression that matches the address, symbol
type and symbol name, and mangles the captured addresses and the names
of symbols of interest into #define directives that are added to
zoffset.h
The symbol type is indicated by a single letter, which we match
strictly: only letters in the set 'ABCDGRSTVW' are matched, even
though the actual symbol type is relevant and therefore ignored.
Commit bc7c9d620 ("efi/libstub/x86: Force 'hidden' visibility for
extern declarations") made a change to the way external symbol
references are classified, resulting in 'startup_32' now being
emitted as a hidden symbol. This prevents the use of GOT entries to
refer to this symbol via its absolute address, which recent toolchains
(including Clang based ones) already avoid by default, making this
change a no-op in the majority of cases.
However, as it turns out, the LLVM linker classifies such hidden
symbols as symbols with static linkage in fully linked ELF binaries,
causing tools such as NM to output a lowercase 't' rather than an upper
case 'T' for the type of such symbols. Since our sed expression only
matches upper case letters for the symbol type, the line describing
startup_32 is disregarded, resulting in a build error like the following
arch/x86/boot/header.S:568:18: error: symbol 'ZO_startup_32' can not be
undefined in a subtraction expression
init_size: .long (0x00000000008fd000 - ZO_startup_32 +
(((0x0000000001f6361c + ((0x0000000001f6361c >> 8) + 65536)
- 0x00000000008c32e5) + 4095) & ~4095)) # kernel initialization size
Given that we are only interested in the value of the symbol, let's match
any character in the set 'a-zA-Z' instead.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Tested-by: Nathan Chancellor <natechancellor@gmail.com>
When installing the EFI virtual address map during early boot, we
access the EFI system table to retrieve the 1:1 mapped address of
the SetVirtualAddressMap() EFI runtime service. This memory is not
known to KASAN, so on KASAN enabled builds, this may result in a
splat like
==================================================================
BUG: KASAN: user-memory-access in efi_set_virtual_address_map+0x141/0x354
Read of size 4 at addr 000000003fbeef38 by task swapper/0/0
CPU: 0 PID: 0 Comm: swapper/0 Not tainted 5.5.0-rc5+ #758
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 0.0.0 02/06/2015
Call Trace:
dump_stack+0x8b/0xbb
? efi_set_virtual_address_map+0x141/0x354
? efi_set_virtual_address_map+0x141/0x354
__kasan_report+0x176/0x192
? efi_set_virtual_address_map+0x141/0x354
kasan_report+0xe/0x20
efi_set_virtual_address_map+0x141/0x354
? efi_thunk_runtime_setup+0x148/0x148
? __inc_numa_state+0x19/0x90
? memcpy+0x34/0x50
efi_enter_virtual_mode+0x5fd/0x67d
start_kernel+0x5cd/0x682
? mem_encrypt_init+0x6/0x6
? x86_family+0x5/0x20
? load_ucode_bsp+0x46/0x154
secondary_startup_64+0xa4/0xb0
==================================================================
Since this code runs only a single time during early boot, let's annotate
it as __no_sanitize_address so KASAN disregards it entirely.
Fixes: 6982947045 ("efi/x86: Split SetVirtualAddresMap() wrappers into ...")
Reported-by: Qian Cai <cai@lca.pw>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
In preparation for fixing efi_memmap_alloc() leaks, add support for
recording whether the memmap was dynamically allocated from slab,
memblock, or is the original physical memmap provided by the platform.
Given this tracking is established in efi_memmap_alloc() and needs to be
carried to efi_memmap_install(), use 'struct efi_memory_map_data' to
convey the flags.
Some small cleanups result from this reorganization, specifically the
removal of local variables for 'phys' and 'size' that are already
tracked in @data.
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20200113172245.27925-12-ardb@kernel.org
We carry a quirk in the x86 EFI code to switch back to an older
method of mapping the EFI runtime services memory regions, because
it was deemed risky at the time to implement a new method without
providing a fallback to the old method in case problems arose.
Such problems did arise, but they appear to be limited to SGI UV1
machines, and so these are the only ones for which the fallback gets
enabled automatically (via a DMI quirk). The fallback can be enabled
manually as well, by passing efi=old_map, but there is very little
evidence that suggests that this is something that is being relied
upon in the field.
Given that UV1 support is not enabled by default by the distros
(Ubuntu, Fedora), there is no point in carrying this fallback code
all the time if there are no other users. So let's move it into the
UV support code, and document that efi=old_map now requires this
support code to be enabled.
Note that efi=old_map has been used in the past on other SGI UV
machines to work around kernel regressions in production, so we
keep the option to enable it by hand, but only if the kernel was
built with UV support.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20200113172245.27925-8-ardb@kernel.org
The EFI code creates RWX mappings for all memory regions that are
occupied after the stub completes, and in the mixed mode case, it
even creates RWX mappings for all of the remaining DRAM as well.
Let's try to avoid this, by setting the NX bit for all memory
regions except the ones that are marked as EFI runtime services
code [which means text+rodata+data in practice, so we cannot mark
them read-only right away]. For cases of buggy firmware where boot
services code is called during SetVirtualAddressMap(), map those
regions with exec permissions as well - they will be unmapped in
efi_free_boot_services().
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20200113172245.27925-7-ardb@kernel.org
The mixed mode thunking routine requires a part of it to be
mapped 1:1, and for this reason, we currently map the entire
kernel .text read/write in the EFI page tables, which is bad.
In fact, the kernel_map_pages_in_pgd() invocation that installs
this mapping is entirely redundant, since all of DRAM is already
1:1 mapped read/write in the EFI page tables when we reach this
point, which means that .rodata is mapped read-write as well.
So let's remap both .text and .rodata read-only in the EFI
page tables.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20200113172245.27925-6-ardb@kernel.org
The following commit:
15f003d207 ("x86/mm/pat: Don't implicitly allow _PAGE_RW in kernel_map_pages_in_pgd()")
modified kernel_map_pages_in_pgd() to manage writable permissions
of memory mappings in the EFI page table in a different way, but
in the process, it removed the ability to clear NX attributes from
read-only mappings, by clobbering the clear mask if _PAGE_RW is not
being requested.
Failure to remove the NX attribute from read-only mappings is
unlikely to be a security issue, but it does prevent us from
tightening the permissions in the EFI page tables going forward,
so let's fix it now.
Fixes: 15f003d207 ("x86/mm/pat: Don't implicitly allow _PAGE_RW in kernel_map_pages_in_pgd()
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20200113172245.27925-5-ardb@kernel.org
The only users of these got removed, so they also need to be
removed to avoid warnings:
arch/x86/boot/compressed/eboot.c: In function 'setup_efi_pci':
arch/x86/boot/compressed/eboot.c:117:16: error: unused variable 'nr_pci' [-Werror=unused-variable]
unsigned long nr_pci;
^~~~~~
arch/x86/boot/compressed/eboot.c: In function 'setup_uga':
arch/x86/boot/compressed/eboot.c:244:16: error: unused variable 'nr_ugas' [-Werror=unused-variable]
unsigned long nr_ugas;
^~~~~~~
Fixes: 2732ea0d5c ("efi/libstub: Use a helper to iterate over a EFI handle array")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20200113172245.27925-4-ardb@kernel.org
Reduce the stack frame of the EFI stub's mixed mode thunk routine by
8 bytes, by moving the GDT and return addresses to EBP and EBX, which
we need to preserve anyway, since their top halves will be cleared by
the call into 32-bit firmware code. Doing so results in the UEFI code
being entered with a 16 byte aligned stack, as mandated by the UEFI
spec, fixing the last occurrence in the 64-bit kernel where we violate
this requirement.
Also, move the saved GDT from a global variable to an unused part of the
stack frame, and touch up some other parts of the code.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20200113172245.27925-3-ardb@kernel.org
Reshuffle the x86 stub code a bit so that we can tag the efi_is_64bit()
function with the 'const' attribute, which permits the compiler to
optimize away any redundant calls. Since we have two different entry
points for 32 and 64 bit firmware in the startup code, this also
simplifies the C code since we'll enter it with the efi_is64 variable
already set.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20200113172245.27925-2-ardb@kernel.org
Pull x86 fixes from Ingo Molnar:
"Misc fixes:
- a resctrl fix for uninitialized objects found by debugobjects
- a resctrl memory leak fix
- fix the unintended re-enabling of the of SME and SEV CPU flags if
memory encryption was disabled at bootup via the MSR space"
* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/CPU/AMD: Ensure clearing of SME/SEV features is maintained
x86/resctrl: Fix potential memory leak
x86/resctrl: Fix an imbalance in domain_remove_cpu()
Pull x86 RAS fix from Ingo Molnar:
"Fix a thermal throttling race that can result in easy to trigger boot
crashes on certain Ice Lake platforms"
* 'ras-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/mce/therm_throt: Do not access uninitialized therm_work
Pull perf fixes from Ingo Molnar:
"Tooling fixes, three Intel uncore driver fixes, plus an AUX events fix
uncovered by the perf fuzzer"
* 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
perf/x86/intel/uncore: Remove PCIe3 unit for SNR
perf/x86/intel/uncore: Fix missing marker for snr_uncore_imc_freerunning_events
perf/x86/intel/uncore: Add PCI ID of IMC for Xeon E3 V5 Family
perf: Correctly handle failed perf_get_aux_event()
perf hists: Fix variable name's inconsistency in hists__for_each() macro
perf map: Set kmap->kmaps backpointer for main kernel map chunks
perf report: Fix incorrectly added dimensions as switch perf data file
tools lib traceevent: Fix memory leakage in filter_event
Pull EFI fixes from Ingo Molnar:
"Three EFI fixes:
- Fix a slow-boot-scrolling regression but making sure we use WC for
EFI earlycon framebuffer mappings on x86
- Fix a mixed EFI mode boot crash
- Disable paging explicitly before entering startup_32() in mixed
mode bootup"
* 'efi-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/efistub: Disable paging at mixed mode entry
efi/libstub/random: Initialize pointer variables to zero for mixed mode
efi/earlycon: Fix write-combine mapping on x86
/* Background. */
For a very long time, extending openat(2) with new features has been
incredibly frustrating. This stems from the fact that openat(2) is
possibly the most famous counter-example to the mantra "don't silently
accept garbage from userspace" -- it doesn't check whether unknown flags
are present[1].
This means that (generally) the addition of new flags to openat(2) has
been fraught with backwards-compatibility issues (O_TMPFILE has to be
defined as __O_TMPFILE|O_DIRECTORY|[O_RDWR or O_WRONLY] to ensure old
kernels gave errors, since it's insecure to silently ignore the
flag[2]). All new security-related flags therefore have a tough road to
being added to openat(2).
Userspace also has a hard time figuring out whether a particular flag is
supported on a particular kernel. While it is now possible with
contemporary kernels (thanks to [3]), older kernels will expose unknown
flag bits through fcntl(F_GETFL). Giving a clear -EINVAL during
openat(2) time matches modern syscall designs and is far more
fool-proof.
In addition, the newly-added path resolution restriction LOOKUP flags
(which we would like to expose to user-space) don't feel related to the
pre-existing O_* flag set -- they affect all components of path lookup.
We'd therefore like to add a new flag argument.
Adding a new syscall allows us to finally fix the flag-ignoring problem,
and we can make it extensible enough so that we will hopefully never
need an openat3(2).
/* Syscall Prototype. */
/*
* open_how is an extensible structure (similar in interface to
* clone3(2) or sched_setattr(2)). The size parameter must be set to
* sizeof(struct open_how), to allow for future extensions. All future
* extensions will be appended to open_how, with their zero value
* acting as a no-op default.
*/
struct open_how { /* ... */ };
int openat2(int dfd, const char *pathname,
struct open_how *how, size_t size);
/* Description. */
The initial version of 'struct open_how' contains the following fields:
flags
Used to specify openat(2)-style flags. However, any unknown flag
bits or otherwise incorrect flag combinations (like O_PATH|O_RDWR)
will result in -EINVAL. In addition, this field is 64-bits wide to
allow for more O_ flags than currently permitted with openat(2).
mode
The file mode for O_CREAT or O_TMPFILE.
Must be set to zero if flags does not contain O_CREAT or O_TMPFILE.
resolve
Restrict path resolution (in contrast to O_* flags they affect all
path components). The current set of flags are as follows (at the
moment, all of the RESOLVE_ flags are implemented as just passing
the corresponding LOOKUP_ flag).
RESOLVE_NO_XDEV => LOOKUP_NO_XDEV
RESOLVE_NO_SYMLINKS => LOOKUP_NO_SYMLINKS
RESOLVE_NO_MAGICLINKS => LOOKUP_NO_MAGICLINKS
RESOLVE_BENEATH => LOOKUP_BENEATH
RESOLVE_IN_ROOT => LOOKUP_IN_ROOT
open_how does not contain an embedded size field, because it is of
little benefit (userspace can figure out the kernel open_how size at
runtime fairly easily without it). It also only contains u64s (even
though ->mode arguably should be a u16) to avoid having padding fields
which are never used in the future.
Note that as a result of the new how->flags handling, O_PATH|O_TMPFILE
is no longer permitted for openat(2). As far as I can tell, this has
always been a bug and appears to not be used by userspace (and I've not
seen any problems on my machines by disallowing it). If it turns out
this breaks something, we can special-case it and only permit it for
openat(2) but not openat2(2).
After input from Florian Weimer, the new open_how and flag definitions
are inside a separate header from uapi/linux/fcntl.h, to avoid problems
that glibc has with importing that header.
/* Testing. */
In a follow-up patch there are over 200 selftests which ensure that this
syscall has the correct semantics and will correctly handle several
attack scenarios.
In addition, I've written a userspace library[4] which provides
convenient wrappers around openat2(RESOLVE_IN_ROOT) (this is necessary
because no other syscalls support RESOLVE_IN_ROOT, and thus lots of care
must be taken when using RESOLVE_IN_ROOT'd file descriptors with other
syscalls). During the development of this patch, I've run numerous
verification tests using libpathrs (showing that the API is reasonably
usable by userspace).
/* Future Work. */
Additional RESOLVE_ flags have been suggested during the review period.
These can be easily implemented separately (such as blocking auto-mount
during resolution).
Furthermore, there are some other proposed changes to the openat(2)
interface (the most obvious example is magic-link hardening[5]) which
would be a good opportunity to add a way for userspace to restrict how
O_PATH file descriptors can be re-opened.
Another possible avenue of future work would be some kind of
CHECK_FIELDS[6] flag which causes the kernel to indicate to userspace
which openat2(2) flags and fields are supported by the current kernel
(to avoid userspace having to go through several guesses to figure it
out).
[1]: https://lwn.net/Articles/588444/
[2]: https://lore.kernel.org/lkml/CA+55aFyyxJL1LyXZeBsf2ypriraj5ut1XkNDsunRBqgVjZU_6Q@mail.gmail.com
[3]: commit 629e014bb8 ("fs: completely ignore unknown open flags")
[4]: https://sourceware.org/bugzilla/show_bug.cgi?id=17523
[5]: https://lore.kernel.org/lkml/20190930183316.10190-2-cyphar@cyphar.com/
[6]: https://youtu.be/ggD-eb3yPVs
Suggested-by: Christian Brauner <christian.brauner@ubuntu.com>
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Currently, there are three static keys in the resctrl file system:
rdt_mon_enable_key and rdt_alloc_enable_key indicate if the monitoring
feature and the allocation feature are enabled, respectively. The
rdt_enable_key is enabled when either the monitoring feature or the
allocation feature is enabled.
If no monitoring feature is present (either hardware doesn't support a
monitoring feature or the feature is disabled by the kernel command line
option "rdt="), rdt_enable_key is still enabled but rdt_mon_enable_key
is disabled.
MBM is a monitoring feature. The MBM overflow handler intends to
check if the monitoring feature is not enabled for fast return.
So check the rdt_mon_enable_key in it instead of the rdt_enable_key as
former is the more accurate check.
[ bp: Massage commit message. ]
Fixes: e33026831b ("x86/intel_rdt/mbm: Handle counter overflow")
Signed-off-by: Xiaochen Shen <xiaochen.shen@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/1576094705-13660-1-git-send-email-xiaochen.shen@intel.com
New Zhaoxin family 7 CPUs are not affected by the SWAPGS vulnerability. So
mark these CPUs in the cpu vulnerability whitelist accordingly.
Signed-off-by: Tony W Wang-oc <TonyWWang-oc@zhaoxin.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/1579227872-26972-3-git-send-email-TonyWWang-oc@zhaoxin.com
New Zhaoxin family 7 CPUs are not affected by SPECTRE_V2. So define a
separate cpu_vuln_whitelist bit NO_SPECTRE_V2 and add these CPUs to the cpu
vulnerability whitelist.
Signed-off-by: Tony W Wang-oc <TonyWWang-oc@zhaoxin.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/1579227872-26972-2-git-send-email-TonyWWang-oc@zhaoxin.com
/proc/cpuinfo currently reports Hardware Lock Elision (HLE) feature to
be present on boot cpu even if it was disabled during the bootup. This
is because cpuinfo_x86->x86_capability HLE bit is not updated after TSX
state is changed via the new MSR IA32_TSX_CTRL.
Update the cached HLE bit also since it is expected to change after an
update to CPUID_CLEAR bit in MSR IA32_TSX_CTRL.
Fixes: 95c5824f75 ("x86/cpu: Add a "tsx=" cmdline option with TSX disabled by default")
Signed-off-by: Pawan Gupta <pawan.kumar.gupta@linux.intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Neelima Krishnan <neelima.krishnan@intel.com>
Reviewed-by: Dave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/2529b99546294c893dfa1c89e2b3e46da3369a59.1578685425.git.pawan.kumar.gupta@linux.intel.com
When CONFIG_PROC_FS is disabled, the compiler warns about an unused
variable:
arch/x86/kernel/apic/x2apic_uv_x.c: In function 'uv_setup_proc_files':
arch/x86/kernel/apic/x2apic_uv_x.c:1546:8: error: unused variable 'name' [-Werror=unused-variable]
char *name = hubless ? "hubless" : "hubbed";
Simplify the code so this variable is no longer needed.
Fixes: 8785968bce ("x86/platform/uv: Add UV Hubbed/Hubless Proc FS Files")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20191212140419.315264-1-arnd@arndb.de
The PCIe Root Port driver for CPU Complex PCIe Root Ports are not
loaded on SNR.
The device ID for SNR PCIe3 unit is used by both uncore driver and the
PCIe Root Port driver. If uncore driver is loaded, the PCIe Root Port
driver never be probed.
Remove the PCIe3 unit for SNR for now. The support for PCIe3 unit will
be added later separately.
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lkml.kernel.org/r/20200116200210.18937-2-kan.liang@linux.intel.com
An Oops during the boot is found on some SNR machines. It turns out
this is because the snr_uncore_imc_freerunning_events[] array was
missing an end-marker.
Fixes: ee49532b38 ("perf/x86/intel/uncore: Add IMC uncore support for Snow Ridge")
Reported-by: Like Xu <like.xu@linux.intel.com>
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Tested-by: Like Xu <like.xu@linux.intel.com>
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20200116200210.18937-1-kan.liang@linux.intel.com
The IMC uncore support is missed for E3-1585 v5 CPU.
Intel Xeon E3 V5 Family has Sky Lake CPU.
Add the PCI ID of IMC for Intel Xeon E3 V5 Family.
Reported-by: Rosales-fernandez, Carlos <carlos.rosales-fernandez@intel.com>
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Tested-by: Rosales-fernandez, Carlos <carlos.rosales-fernandez@intel.com>
Link: https://lkml.kernel.org/r/1578687311-158748-1-git-send-email-kan.liang@linux.intel.com
Description of hardware operation
---------------------------------
The core AMD PMU has a 4-bit wide per-cycle increment for each
performance monitor counter. That works for most events, but
now with AMD Family 17h and above processors, some events can
occur more than 15 times in a cycle. Those events are called
"Large Increment per Cycle" events. In order to count these
events, two adjacent h/w PMCs get their count signals merged
to form 8 bits per cycle total. In addition, the PERF_CTR count
registers are merged to be able to count up to 64 bits.
Normally, events like instructions retired, get programmed on a single
counter like so:
PERF_CTL0 (MSR 0xc0010200) 0x000000000053ff0c # event 0x0c, umask 0xff
PERF_CTR0 (MSR 0xc0010201) 0x0000800000000001 # r/w 48-bit count
The next counter at MSRs 0xc0010202-3 remains unused, or can be used
independently to count something else.
When counting Large Increment per Cycle events, such as FLOPs,
however, we now have to reserve the next counter and program the
PERF_CTL (config) register with the Merge event (0xFFF), like so:
PERF_CTL0 (msr 0xc0010200) 0x000000000053ff03 # FLOPs event, umask 0xff
PERF_CTR0 (msr 0xc0010201) 0x0000800000000001 # rd 64-bit cnt, wr lo 48b
PERF_CTL1 (msr 0xc0010202) 0x0000000f004000ff # Merge event, enable bit
PERF_CTR1 (msr 0xc0010203) 0x0000000000000000 # wr hi 16-bits count
The count is widened from the normal 48-bits to 64 bits by having the
second counter carry the higher 16 bits of the count in its lower 16
bits of its counter register.
The odd counter, e.g., PERF_CTL1, is programmed with the enabled Merge
event before the even counter, PERF_CTL0.
The Large Increment feature is available starting with Family 17h.
For more details, search any Family 17h PPR for the "Large Increment
per Cycle Events" section, e.g., section 2.1.15.3 on p. 173 in this
version:
https://www.amd.com/system/files/TechDocs/56176_ppr_Family_17h_Model_71h_B0_pub_Rev_3.06.zip
Description of software operation
---------------------------------
The following steps are taken in order to support reserving and
enabling the extra counter for Large Increment per Cycle events:
1. In the main x86 scheduler, we reduce the number of available
counters by the number of Large Increment per Cycle events being
scheduled, tracked by a new cpuc variable 'n_pair' and a new
amd_put_event_constraints_f17h(). This improves the counter
scheduler success rate.
2. In perf_assign_events(), if a counter is assigned to a Large
Increment event, we increment the current counter variable, so the
counter used for the Merge event is removed from assignment
consideration by upcoming event assignments.
3. In find_counter(), if a counter has been found for the Large
Increment event, we set the next counter as used, to prevent other
events from using it.
4. We perform steps 2 & 3 also in the x86 scheduler fastpath, i.e.,
we add Merge event accounting to the existing used_mask logic.
5. Finally, we add on the programming of Merge event to the
neighbouring PMC counters in the counter enable/disable{_all}
code paths.
Currently, software does not support a single PMU with mixed 48- and
64-bit counting, so Large increment event counts are limited to 48
bits. In set_period, we zero-out the upper 16 bits of the count, so
the hardware doesn't copy them to the even counter's higher bits.
Simple invocation example showing counting 8 FLOPs per 256-bit/%ymm
vaddps instruction executed in a loop 100 million times:
perf stat -e cpu/fp_ret_sse_avx_ops.all/,cpu/instructions/ <workload>
Performance counter stats for '<workload>':
800,000,000 cpu/fp_ret_sse_avx_ops.all/u
300,042,101 cpu/instructions/u
Prior to this patch, the reported SSE/AVX FLOPs retired count would
be wrong.
[peterz: lots of renames and edits to the code]
Signed-off-by: Kim Phillips <kim.phillips@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
AMD Family 17h processors and above gain support for Large Increment
per Cycle events. Unfortunately there is no CPUID or equivalent bit
that indicates whether the feature exists or not, so we continue to
determine eligibility based on a CPU family number comparison.
For Large Increment per Cycle events, we add a f17h-and-compatibles
get_event_constraints_f17h() that returns an even counter bitmask:
Large Increment per Cycle events can only be placed on PMCs 0, 2,
and 4 out of the currently available 0-5. The only currently
public event that requires this feature to report valid counts
is PMCx003 "Retired SSE/AVX Operations".
Note that the CPU family logic in amd_core_pmu_init() is changed
so as to be able to selectively add initialization for features
available in ranges of backward-compatible CPU families. This
Large Increment per Cycle feature is expected to be retained
in future families.
A side-effect of assigning a new get_constraints function for f17h
disables calling the old (prior to f15h) amd_get_event_constraints
implementation left enabled by commit e40ed1542d ("perf/x86: Add perf
support for AMD family-17h processors"), which is no longer
necessary since those North Bridge event codes are obsoleted.
Also fix a spelling mistake whilst in the area (calulating ->
calculating).
Fixes: e40ed1542d ("perf/x86: Add perf support for AMD family-17h processors")
Signed-off-by: Kim Phillips <kim.phillips@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20191114183720.19887-2-kim.phillips@amd.com
If the SME and SEV features are present via CPUID, but memory encryption
support is not enabled (MSR 0xC001_0010[23]), the feature flags are cleared
using clear_cpu_cap(). However, if get_cpu_cap() is later called, these
feature flags will be reset back to present, which is not desired.
Change from using clear_cpu_cap() to setup_clear_cpu_cap() so that the
clearing of the flags is maintained.
Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: <stable@vger.kernel.org> # 4.16.x-
Link: https://lkml.kernel.org/r/226de90a703c3c0be5a49565047905ac4e94e8f3.1579125915.git.thomas.lendacky@amd.com
Add the new PCI Device 18h IDs for AMD Family 19h systems. Note that
Family 19h systems will not have a new PCI root device ID.
Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20200110015651.14887-4-Yazen.Ghannam@amd.com
Add support for a new version of the Load Store unit bank type as
indicated by its McaType value, which will be present in future SMCA
systems.
Add the new (HWID, MCATYPE) tuple. Reuse the same name, since this is
logically the same to the user.
Also, add the new error descriptions to edac_mce_amd.
Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20200110015651.14887-2-Yazen.Ghannam@amd.com
These x86_64 vectorized implementations come from Andy Polyakov's
CRYPTOGAMS implementation, and are included here in raw form without
modification, so that subsequent commits that fix these up for the
kernel can see how it has changed.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
These two C implementations from Zinc -- a 32x32 one and a 64x64 one,
depending on the platform -- come from Andrew Moon's public domain
poly1305-donna portable code, modified for usage in the kernel. The
precomputation in the 32-bit version and the use of 64x64 multiplies in
the 64-bit version make these perform better than the code it replaces.
Moon's code is also very widespread and has received many eyeballs of
scrutiny.
There's a bit of interference between the x86 implementation, which
relies on internal details of the old scalar implementation. In the next
commit, the x86 implementation will be replaced with a faster one that
doesn't rely on this, so none of this matters much. But for now, to keep
this passing the tests, we inline the bits of the old implementation
that the x86 implementation relied on. Also, since we now support a
slightly larger key space, via the union, some offsets had to be fixed
up.
Nonce calculation was folded in with the emit function, to take
advantage of 64x64 arithmetic. However, Adiantum appeared to rely on no
nonce handling in emit, so this path was conditionalized. We also
introduced a new struct, poly1305_core_key, to represent the precise
amount of space that particular implementation uses.
Testing with kbench9000, depending on the CPU, the update function for
the 32x32 version has been improved by 4%-7%, and for the 64x64 by
19%-30%. The 32x32 gains are small, but I think there's great value in
having a parallel implementation to the 64x64 one so that the two can be
compared side-by-side as nice stand-alone units.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Don't print an error message about VMX being disabled by BIOS if KVM,
the sole user of VMX, is disabled. E.g. if KVM is disabled and the MSR
is unlocked, the kernel will intentionally disable VMX when locking
feature control and then complain that "BIOS" disabled VMX.
Fixes: ef4d3bf198 ("x86/cpu: Clear VMX feature flag if VMX is not fully enabled")
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20200114202545.20296-1-sean.j.christopherson@intel.com
It is relatively easy to trigger the following boot splat on an Ice Lake
client platform. The call stack is like:
kernel BUG at kernel/timer/timer.c:1152!
Call Trace:
__queue_delayed_work
queue_delayed_work_on
therm_throt_process
intel_thermal_interrupt
...
The reason is that a CPU's thermal interrupt is enabled prior to
executing its hotplug onlining callback which will initialize the
throttling workqueues.
Such a race can lead to therm_throt_process() accessing an uninitialized
therm_work, leading to the above BUG at a very early bootup stage.
Therefore, unmask the thermal interrupt vector only after having setup
the workqueues completely.
[ bp: Heavily massage commit message and correct comment formatting. ]
Fixes: f6656208f0 ("x86/mce/therm_throt: Optimize notifications of thermal throttle")
Signed-off-by: Chuansheng Liu <chuansheng.liu@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Acked-by: Tony Luck <tony.luck@intel.com>
Link: https://lkml.kernel.org/r/20200107004116.59353-1-chuansheng.liu@intel.com
con_init in tty/vt.c will now set conswitchp to dummy_con if it's unset.
Drop it from arch setup code.
Signed-off-by: Arvind Sankar <nivedita@alum.mit.edu>
Link: https://lore.kernel.org/r/20191218214506.49252-24-nivedita@alum.mit.edu
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The VVAR page layout depends on whether a task belongs to the root or
non-root time namespace. Whenever a task changes its namespace, the VVAR
page tables are cleared and then they will be re-faulted with a
corresponding layout.
Co-developed-by: Andrei Vagin <avagin@gmail.com>
Signed-off-by: Andrei Vagin <avagin@gmail.com>
Signed-off-by: Dmitry Safonov <dima@arista.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20191112012724.250792-27-dima@arista.com
As timens page has offsets to data on VVAR page VVAR is going
to be accessed shortly. Set it up with timens in one page fault
as optimization.
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Co-developed-by: Andrei Vagin <avagin@gmail.com>
Signed-off-by: Andrei Vagin <avagin@gmail.com>
Signed-off-by: Dmitry Safonov <dima@arista.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20191112012724.250792-26-dima@arista.com
If a task belongs to a time namespace then the VVAR page which contains
the system wide VDSO data is replaced with a namespace specific page
which has the same layout as the VVAR page.
Co-developed-by: Andrei Vagin <avagin@gmail.com>
Signed-off-by: Andrei Vagin <avagin@gmail.com>
Signed-off-by: Dmitry Safonov <dima@arista.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20191112012724.250792-25-dima@arista.com
To support time namespaces in the VDSO with a minimal impact on regular non
time namespace affected tasks, the namespace handling needs to be hidden in
a slow path.
The most obvious place is vdso_seq_begin(). If a task belongs to a time
namespace then the VVAR page which contains the system wide VDSO data is
replaced with a namespace specific page which has the same layout as the
VVAR page. That page has vdso_data->seq set to 1 to enforce the slow path
and vdso_data->clock_mode set to VCLOCK_TIMENS to enforce the time
namespace handling path.
The extra check in the case that vdso_data->seq is odd, e.g. a concurrent
update of the VDSO data is in progress, is not really affecting regular
tasks which are not part of a time namespace as the task is spin waiting
for the update to finish and vdso_data->seq to become even again.
If a time namespace task hits that code path, it invokes the corresponding
time getter function which retrieves the real VVAR page, reads host time
and then adds the offset for the requested clock which is stored in the
special VVAR page.
Allocate the time namespace page among VVAR pages and place vdso_data on
it. Provide __arch_get_timens_vdso_data() helper for VDSO code to get the
code-relative position of VVARs on that special page.
Co-developed-by: Andrei Vagin <avagin@openvz.org>
Signed-off-by: Andrei Vagin <avagin@openvz.org>
Signed-off-by: Dmitry Safonov <dima@arista.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20191112012724.250792-23-dima@arista.com
VDSO support for time namespaces needs to set up a page with the same
layout as VVAR. That timens page will be placed on position of VVAR page
inside namespace. That page has vdso_data->seq set to 1 to enforce
the slow path and vdso_data->clock_mode set to VCLOCK_TIMENS to enforce
the time namespace handling path.
To prepare the time namespace page the kernel needs to know the vdso_data
offset. Provide arch_get_vdso_data() helper for locating vdso_data on VVAR
page.
Co-developed-by: Andrei Vagin <avagin@openvz.org>
Signed-off-by: Andrei Vagin <avagin@openvz.org>
Signed-off-by: Dmitry Safonov <dima@arista.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20191112012724.250792-22-dima@arista.com
Forbid splitting VVAR VMA resulting in a stricter ABI and reducing the
amount of corner-cases to consider while working further on VDSO time
namespace support.
As the offset from timens to VVAR page is computed compile-time, the pages
in VVAR should stay together and not being partically mremap()'ed.
Co-developed-by: Andrei Vagin <avagin@openvz.org>
Signed-off-by: Andrei Vagin <avagin@openvz.org>
Signed-off-by: Dmitry Safonov <dima@arista.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20191112012724.250792-20-dima@arista.com
VDSO_HAS_32BIT_FALLBACK has been removed from the core since
the architectures that support the generic vDSO library have
been converted to support the 32 bit fallbacks.
Remove unused VDSO_HAS_32BIT_FALLBACK from x86 vdso.
Signed-off-by: Vincenzo Frascino <vincenzo.frascino@arm.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20190830135902.20861-9-vincenzo.frascino@arm.com
This wires up the pidfd_getfd syscall for all architectures.
Signed-off-by: Sargun Dhillon <sargun@sargun.me>
Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Arnd Bergmann <arnd@arndb.de>
Link: https://lore.kernel.org/r/20200107175927.4558-4-sargun@sargun.me
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Change the dependency for KVM_INTEL, i.e. KVM w/ VMX, from Intel CPUs to
any CPU that supports the IA32_FEAT_CTL MSR and thus VMX functionality.
This effectively allows building KVM_INTEL for Centaur and Zhaoxin CPUs.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20191221044513.21680-20-sean.j.christopherson@intel.com
Provide stubs for perf_guest_get_msrs() and intel_pt_handle_vmx() when
building without support for Intel CPUs, i.e. CPU_SUP_INTEL=n. Lack of
stubs is not currently a problem as the only user, KVM_INTEL, takes a
dependency on CPU_SUP_INTEL=y. Provide the stubs for all CPUs so that
KVM_INTEL can be built for any CPU with compatible hardware support,
e.g. Centuar and Zhaoxin CPUs.
Note, the existing stub for perf_guest_get_msrs() is essentially dead
code as KVM selects CONFIG_PERF_EVENTS, i.e. the only user guarantees
the full implementation is built.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20191221044513.21680-19-sean.j.christopherson@intel.com
Define the VMCS execution control flags (consumed by KVM) using their
associated VMX_FEATURE_* to provide a strong hint that new VMX features
are expected to be added to VMX_FEATURE and considered for reporting via
/proc/cpuinfo.
No functional change intended.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20191221044513.21680-18-sean.j.christopherson@intel.com
Explicitly check the current CPU's IA32_FEAT_CTL and VMX feature flags
when verifying compatibility across physical CPUs. This effectively
adds a check on IA32_FEAT_CTL to ensure that VMX is fully enabled on
all CPUs.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20191221044513.21680-17-sean.j.christopherson@intel.com
Replace KVM's manual checks on IA32_FEAT_CTL with a query on the boot
CPU's MSR_IA32_FEAT_CTL and VMX feature flags. The MSR_IA32_FEAT_CTL
indicates that IA32_FEAT_CTL has been configured and that dependent
features are accurately reflected in cpufeatures, e.g. the VMX flag is
now cleared during boot if VMX isn't fully enabled via IA32_FEAT_CTL,
including the case where the MSR isn't supported.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20191221044513.21680-16-sean.j.christopherson@intel.com
Remove KVM's code to initialize IA32_FEAT_CTL MSR when KVM is loaded now
that the MSR is initialized during boot on all CPUs that support VMX,
i.e. on all CPUs that can possibly load kvm_intel.
Note, don't WARN if IA32_FEAT_CTL is unlocked, even though the MSR is
unconditionally locked by init_ia32_feat_ctl(). KVM isn't tied directly
to a CPU vendor detection, whereas init_ia32_feat_ctl() is invoked if
and only if the CPU vendor is recognized and known to support VMX. As a
result, vmx_disabled_by_bios() may be reached without going through
init_ia32_feat_ctl() and thus without locking IA32_FEAT_CTL. This quirk
will be eliminated in a future patch.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Reviewed-by: Jim Mattson <jmattson@google.com>
Link: https://lkml.kernel.org/r/20191221044513.21680-15-sean.j.christopherson@intel.com
Add a new feature flag, X86_FEATURE_MSR_IA32_FEAT_CTL, to track whether
IA32_FEAT_CTL has been initialized. This will allow KVM, and any future
subsystems that depend on IA32_FEAT_CTL, to rely purely on cpufeatures
to query platform support, e.g. allows a future patch to remove KVM's
manual IA32_FEAT_CTL MSR checks.
Various features (on platforms that support IA32_FEAT_CTL) are dependent
on IA32_FEAT_CTL being configured and locked, e.g. VMX and LMCE. The
MSR is always configured during boot, but only if the CPU vendor is
recognized by the kernel. Because CPUID doesn't incorporate the current
IA32_FEAT_CTL value in its reporting of relevant features, it's possible
for a feature to be reported as supported in cpufeatures but not truly
enabled, e.g. if the CPU supports VMX but the kernel doesn't recognize
the CPU.
As a result, without the flag, KVM would see VMX as supported even if
IA32_FEAT_CTL hasn't been initialized, and so would need to manually
read the MSR and check the various enabling bits to avoid taking an
unexpected #GP on VMXON.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20191221044513.21680-14-sean.j.christopherson@intel.com
Set the synthetic VMX cpufeatures, which need to be kept to preserve
/proc/cpuinfo's ABI, in the common IA32_FEAT_CTL initialization code.
Remove the vendor code that manually sets the synthetic flags.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20191221044513.21680-13-sean.j.christopherson@intel.com
Add support for generating VMX feature names in capflags.c and use the
resulting x86_vmx_flags to print the VMX flags in /proc/cpuinfo. Don't
print VMX flags if no bits are set in word 0, which holds Pin Controls.
Pin Control's INTR and NMI exiting are fundamental pillars of VMX, if
they are not supported then the CPU is broken, it does not actually
support VMX, or the kernel wasn't built with support for the target CPU.
Print the features in a dedicated "vmx flags" line to avoid polluting
the common "flags" and to avoid having to prefix all flags with "vmx_",
which results in horrendously long names.
Keep synthetic VMX flags in cpufeatures to preserve /proc/cpuinfo's ABI
for those flags. This means that "flags" and "vmx flags" will have
duplicate entries for tpr_shadow (virtual_tpr), vnmi, ept, flexpriority,
vpid and ept_ad, but caps the pollution of "flags" at those six VMX
features. The vendor-specific code that populates the synthetic flags
will be consolidated in a future patch to further minimize the lasting
damage.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20191221044513.21680-12-sean.j.christopherson@intel.com
Add an entry in struct cpuinfo_x86 to track VMX capabilities and fill
the capabilities during IA32_FEAT_CTL MSR initialization.
Make the VMX capabilities dependent on IA32_FEAT_CTL and
X86_FEATURE_NAMES so as to avoid unnecessary overhead on CPUs that can't
possibly support VMX, or when /proc/cpuinfo is not available.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20191221044513.21680-11-sean.j.christopherson@intel.com
Add a VMX-specific variant of X86_FEATURE_* flags, which will eventually
supplant the synthetic VMX flags defined in cpufeatures word 8. Use the
Intel-defined layouts for the major VMX execution controls so that their
word entries can be directly populated from their respective MSRs, and
so that the VMX_FEATURE_* flags can be used to define the existing bit
definitions in asm/vmx.h, i.e. force developers to define a VMX_FEATURE
flag when adding support for a new hardware feature.
The majority of Intel's (and compatible CPU's) VMX capabilities are
enumerated via MSRs and not CPUID, i.e. querying /proc/cpuinfo doesn't
naturally provide any insight into the virtualization capabilities of
VMX enabled CPUs. Commit
e38e05a858 ("x86: extended "flags" to show virtualization HW feature
in /proc/cpuinfo")
attempted to address the issue by synthesizing select VMX features into
a Linux-defined word in cpufeatures.
Lack of reporting of VMX capabilities via /proc/cpuinfo is problematic
because there is no sane way for a user to query the capabilities of
their platform, e.g. when trying to find a platform to test a feature or
debug an issue that has a hardware dependency. Lack of reporting is
especially problematic when the user isn't familiar with VMX, e.g. the
format of the MSRs is non-standard, existence of some MSRs is reported
by bits in other MSRs, several "features" from KVM's point of view are
enumerated as 3+ distinct features by hardware, etc...
The synthetic cpufeatures approach has several flaws:
- The set of synthesized VMX flags has become extremely stale with
respect to the full set of VMX features, e.g. only one new flag
(EPT A/D) has been added in the the decade since the introduction of
the synthetic VMX features. Failure to keep the VMX flags up to
date is likely due to the lack of a mechanism that forces developers
to consider whether or not a new feature is worth reporting.
- The synthetic flags may incorrectly be misinterpreted as affecting
kernel behavior, i.e. KVM, the kernel's sole consumer of VMX,
completely ignores the synthetic flags.
- New CPU vendors that support VMX have duplicated the hideous code
that propagates VMX features from MSRs to cpufeatures. Bringing the
synthetic VMX flags up to date would exacerbate the copy+paste
trainwreck.
Define separate VMX_FEATURE flags to set the stage for enumerating VMX
capabilities outside of the cpu_has() framework, and for adding
functional usage of VMX_FEATURE_* to help ensure the features reported
via /proc/cpuinfo is up to date with respect to kernel recognition of
VMX capabilities.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20191221044513.21680-10-sean.j.christopherson@intel.com
Now that IA32_FEAT_CTL is always configured and locked for CPUs that are
known to support VMX[*], clear the VMX capability flag if the MSR is
unsupported or BIOS disabled VMX, i.e. locked IA32_FEAT_CTL and didn't
set the appropriate VMX enable bit.
[*] Because init_ia32_feat_ctl() is called from vendors ->c_init(), it's
still possible for IA32_FEAT_CTL to be left unlocked when VMX is
supported by the CPU. This is not fatal, and will be addressed in a
future patch.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20191221044513.21680-9-sean.j.christopherson@intel.com
Use the recently added IA32_FEAT_CTL MSR initialization sequence to
opportunistically enable VMX support when running on a Zhaoxin CPU.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20191221044513.21680-8-sean.j.christopherson@intel.com
Use the recently added IA32_FEAT_CTL MSR initialization sequence to
opportunistically enable VMX support when running on a Centaur CPU.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20191221044513.21680-7-sean.j.christopherson@intel.com
Opportunistically initialize IA32_FEAT_CTL to enable VMX when the MSR is
left unlocked by BIOS. Configuring feature control at boot time paves
the way for similar enabling of other features, e.g. Software Guard
Extensions (SGX).
Temporarily leave equivalent KVM code in place in order to avoid
introducing a regression on Centaur and Zhaoxin CPUs, e.g. removing
KVM's code would leave the MSR unlocked on those CPUs and would break
existing functionality if people are loading kvm_intel on Centaur and/or
Zhaoxin. Defer enablement of the boot-time configuration on Centaur and
Zhaoxin to future patches to aid bisection.
Note, Local Machine Check Exceptions (LMCE) are also supported by the
kernel and enabled via feature control, but the kernel currently uses
LMCE if and only if the feature is explicitly enabled by BIOS. Keep
the current behavior to avoid introducing bugs, future patches can opt
in to opportunistic enabling if it's deemed desirable to do so.
Always lock IA32_FEAT_CTL if it exists, even if the CPU doesn't support
VMX, so that other existing and future kernel code that queries the MSR
can assume it's locked.
Start from a clean slate when constructing the value to write to
IA32_FEAT_CTL, i.e. ignore whatever value BIOS left in the MSR so as not
to enable random features or fault on the WRMSR.
Suggested-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20191221044513.21680-5-sean.j.christopherson@intel.com
As pointed out by Boris, the defines for bits in IA32_FEATURE_CONTROL
are quite a mouthful, especially the VMX bits which must differentiate
between enabling VMX inside and outside SMX (TXT) operation. Rename the
MSR and its bit defines to abbreviate FEATURE_CONTROL as FEAT_CTL to
make them a little friendlier on the eyes.
Arguably, the MSR itself should keep the full IA32_FEATURE_CONTROL name
to match Intel's SDM, but a future patch will add a dedicated Kconfig,
file and functions for the MSR. Using the full name for those assets is
rather unwieldy, so bite the bullet and use IA32_FEAT_CTL so that its
nomenclature is consistent throughout the kernel.
Opportunistically, fix a few other annoyances with the defines:
- Relocate the bit defines so that they immediately follow the MSR
define, e.g. aren't mistaken as belonging to MISC_FEATURE_CONTROL.
- Add whitespace around the block of feature control defines to make
it clear they're all related.
- Use BIT() instead of manually encoding the bit shift.
- Use "VMX" instead of "VMXON" to match the SDM.
- Append "_ENABLED" to the LMCE (Local Machine Check Exception) bit to
be consistent with the kernel's verbiage used for all other feature
control bits. Note, the SDM refers to the LMCE bit as LMCE_ON,
likely to differentiate it from IA32_MCG_EXT_CTL.LMCE_EN. Ignore
the (literal) one-off usage of _ON, the SDM is simply "wrong".
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20191221044513.21680-2-sean.j.christopherson@intel.com
When writing a pid to file "tasks", a callback function move_myself() is
queued to this task to be called when the task returns from kernel mode
or exits. The purpose of move_myself() is to activate the newly assigned
closid and/or rmid associated with this task. This activation is done by
calling resctrl_sched_in() from move_myself(), the same function that is
called when switching to this task.
If this work is successfully queued but then the task enters PF_EXITING
status (e.g., receiving signal SIGKILL, SIGTERM) prior to the
execution of the callback move_myself(), move_myself() still calls
resctrl_sched_in() since the task status is not currently considered.
When a task is exiting, the data structure of the task itself will
be freed soon. Calling resctrl_sched_in() to write the register that
controls the task's resources is unnecessary and it implies extra
performance overhead.
Add check on task status in move_myself() and return immediately if the
task is PF_EXITING.
[ bp: Massage. ]
Signed-off-by: Xiaochen Shen <xiaochen.shen@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Reviewed-by: Reinette Chatre <reinette.chatre@intel.com>
Link: https://lkml.kernel.org/r/1578500026-21152-1-git-send-email-xiaochen.shen@intel.com
The function mce_severity() is not required to update its msg argument.
In fact, mce_severity_amd() does not, which makes mce_no_way_out()
return uninitialized data, which may be used later for printing.
Assuming that implementations of mce_severity() either always or never
update the msg argument (which is currently the case), it is sufficient
to initialize the temporary variable in mce_no_way_out().
While at it, avoid printing a useless "Unknown".
Signed-off-by: Jan H. Schönherr <jschoenh@amazon.de>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20200103150722.20313-4-jschoenh@amazon.de
Since commit
8b38937b7a ("x86/mce: Do not enter deferred errors into the generic
pool twice")
the mce=nobootlog option has become mostly ineffective (after being only
slightly ineffective before), as the code is taking actions on MCEs left
over from boot when they have a usable address.
Move the check for MCP_DONTLOG a bit outward to make it effective again.
Also, since commit
011d826111 ("RAS: Add a Corrected Errors Collector")
the two branches of the remaining "if" at the bottom of machine_check_poll()
do same. Unify them.
Signed-off-by: Jan H. Schönherr <jschoenh@amazon.de>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20200103150722.20313-3-jschoenh@amazon.de
Commit
fa92c58694 ("x86, mce: Support memory error recovery for both UCNA
and Deferred error in machine_check_poll")
added handling of UCNA and Deferred errors by adding them to the ring
for SRAO errors.
Later, commit
fd4cf79fcc ("x86/mce: Remove the MCE ring for Action Optional errors")
switched storage from the SRAO ring to the unified pool that is still
in use today. In order to only act on the intended errors, a filter
for MCE_AO_SEVERITY is used -- effectively removing handling of
UCNA/Deferred errors again.
Extend the severity filter to include UCNA/Deferred errors again.
Also, generalize the naming of the notifier from SRAO to UC to capture
the extended scope.
Note, that this change may cause a message like the following to appear,
as the same address may be reported as SRAO and as UCNA:
Memory failure: 0x5fe3284: already hardware poisoned
Technically, this is a return to previous behavior.
Signed-off-by: Jan H. Schönherr <jschoenh@amazon.de>
Signed-off-by: Borislav Petkov <bp@suse.de>
Acked-by: Tony Luck <tony.luck@intel.com>
Link: https://lkml.kernel.org/r/20200103150722.20313-2-jschoenh@amazon.de
First, printk() is NMI-context safe now since the safe printk() has been
implemented and it already has an irq_work to make NMI-context safe.
Second, this NMI irq_work actually does not work if a NMI handler causes
panic by watchdog timeout. It has no chance to run in such case, while
the safe printk() will flush its per-cpu buffers before panicking.
While at it, repurpose the irq_work callback into a function which
concentrates the NMI duration checking and makes the code easier to
follow.
[ bp: Massage. ]
Signed-off-by: Changbin Du <changbin.du@gmail.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lkml.kernel.org/r/20200111125427.15662-1-changbin.du@gmail.com
Add an option to disable the busmaster bit in the control register on
all PCI bridges before calling ExitBootServices() and passing control
to the runtime kernel. System firmware may configure the IOMMU to prevent
malicious PCI devices from being able to attack the OS via DMA. However,
since firmware can't guarantee that the OS is IOMMU-aware, it will tear
down IOMMU configuration when ExitBootServices() is called. This leaves
a window between where a hostile device could still cause damage before
Linux configures the IOMMU again.
If CONFIG_EFI_DISABLE_PCI_DMA is enabled or "efi=disable_early_pci_dma"
is passed on the command line, the EFI stub will clear the busmaster bit
on all PCI bridges before ExitBootServices() is called. This will
prevent any malicious PCI devices from being able to perform DMA until
the kernel reenables busmastering after configuring the IOMMU.
This option may cause failures with some poorly behaved hardware and
should not be enabled without testing. The kernel commandline options
"efi=disable_early_pci_dma" or "efi=no_disable_early_pci_dma" may be
used to override the default. Note that PCI devices downstream from PCI
bridges are disconnected from their drivers first, using the UEFI
driver model API, so that DMA can be disabled safely at the bridge
level.
[ardb: disconnect PCI I/O handles first, as suggested by Arvind]
Co-developed-by: Matthew Garrett <mjg59@google.com>
Signed-off-by: Matthew Garrett <mjg59@google.com>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Arvind Sankar <nivedita@alum.mit.edu>
Cc: Matthew Garrett <matthewgarrett@google.com>
Cc: linux-efi@vger.kernel.org
Link: https://lkml.kernel.org/r/20200103113953.9571-18-ardb@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Introduce the ability to define macros to perform argument translation
for the calls that need it, and define them for the boot services that
we currently use.
When calling 32-bit firmware methods in mixed mode, all output
parameters that are 32-bit according to the firmware, but 64-bit in the
kernel (ie OUT UINTN * or OUT VOID **) must be initialized in the
kernel, or the upper 32 bits may contain garbage. Define macros that
zero out the upper 32 bits of the output before invoking the firmware
method.
When a 32-bit EFI call takes 64-bit arguments, the mixed-mode call must
push the two 32-bit halves as separate arguments onto the stack. This
can be achieved by splitting the argument into its two halves when
calling the assembler thunk. Define a macro to do this for the
free_pages boot service.
Signed-off-by: Arvind Sankar <nivedita@alum.mit.edu>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Matthew Garrett <mjg59@google.com>
Cc: linux-efi@vger.kernel.org
Link: https://lkml.kernel.org/r/20200103113953.9571-17-ardb@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
On x86 we need to thunk through assembler stubs to call the EFI services
for mixed mode, and for runtime services in 64-bit mode. The assembler
stubs have limits on how many arguments it handles. Introduce a few
macros to check that we do not try to pass too many arguments to the
stubs.
Signed-off-by: Arvind Sankar <nivedita@alum.mit.edu>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Matthew Garrett <mjg59@google.com>
Cc: linux-efi@vger.kernel.org
Link: https://lkml.kernel.org/r/20200103113953.9571-16-ardb@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Remove some code that is guaranteed to be unreachable, given
that we have already bailed by this time if EFI_OLD_MEMMAP is
set.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Arvind Sankar <nivedita@alum.mit.edu>
Cc: Matthew Garrett <mjg59@google.com>
Cc: linux-efi@vger.kernel.org
Link: https://lkml.kernel.org/r/20200103113953.9571-15-ardb@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The logic in __efi_enter_virtual_mode() does a number of steps in
sequence, all of which may fail in one way or the other. In most
cases, we simply print an error and disable EFI runtime services
support, but in some cases, we BUG() or panic() and bring down the
system when encountering conditions that we could easily handle in
the same way.
While at it, replace a pointless page-to-virt-phys conversion with
one that goes straight from struct page to physical.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Arvind Sankar <nivedita@alum.mit.edu>
Cc: Matthew Garrett <mjg59@google.com>
Cc: linux-efi@vger.kernel.org
Link: https://lkml.kernel.org/r/20200103113953.9571-14-ardb@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Clean up the efi_systab_init() routine which maps the EFI system
table and copies the relevant pieces of data out of it.
The current routine is very difficult to read, so let's clean that
up. Also, switch to a R/O mapping of the system table since that is
all we need.
Finally, use a plain u64 variable to record the physical address of
the system table instead of pointlessly stashing it in a struct efi
that is never used for anything else.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Arvind Sankar <nivedita@alum.mit.edu>
Cc: Matthew Garrett <mjg59@google.com>
Cc: linux-efi@vger.kernel.org
Link: https://lkml.kernel.org/r/20200103113953.9571-13-ardb@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The routines efi_runtime_init32() and efi_runtime_init64() are
almost indistinguishable, and the only relevant difference is
the offset in the runtime struct from where to obtain the physical
address of the SetVirtualAddressMap() routine.
However, this address is only used once, when installing the virtual
address map that the OS will use to invoke EFI runtime services, and
at the time of the call, we will necessarily be running with a 1:1
mapping, and so there is no need to do the map/unmap dance here to
retrieve the address. In fact, in the preceding changes to these users,
we stopped using the address recorded here entirely.
So let's just get rid of all this code since it no longer serves a
purpose. While at it, tweak the logic so that we handle unsupported
and disable EFI runtime services in the same way, and unmap the EFI
memory map in both cases.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Arvind Sankar <nivedita@alum.mit.edu>
Cc: Matthew Garrett <mjg59@google.com>
Cc: linux-efi@vger.kernel.org
Link: https://lkml.kernel.org/r/20200103113953.9571-12-ardb@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Calling 32-bit EFI runtime services from a 64-bit OS involves
switching back to the flat mapping with a stack carved out of
memory that is 32-bit addressable.
There is no need to actually execute the 64-bit part of this
routine from the flat mapping as well, as long as the entry
and return address fit in 32 bits. There is also no need to
preserve part of the calling context in global variables: we
can simply push the old stack pointer value to the new stack,
and keep the return address from the code32 section in EBX.
While at it, move the conditional check whether to invoke
the mixed mode version of SetVirtualAddressMap() into the
64-bit implementation of the wrapper routine.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Arvind Sankar <nivedita@alum.mit.edu>
Cc: Matthew Garrett <mjg59@google.com>
Cc: linux-efi@vger.kernel.org
Link: https://lkml.kernel.org/r/20200103113953.9571-11-ardb@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The efi_call() wrapper used to invoke EFI runtime services serves
a number of purposes:
- realign the stack to 16 bytes
- preserve FP and CR0 register state
- translate from SysV to MS calling convention.
Preserving CR0.TS is no longer necessary in Linux, and preserving the
FP register state is also redundant in most cases, since efi_call() is
almost always used from within the scope of a pair of kernel_fpu_begin()/
kernel_fpu_end() calls, with the exception of the early call to
SetVirtualAddressMap() and the SGI UV support code.
So let's add a pair of kernel_fpu_begin()/_end() calls there as well,
and remove the unnecessary code from the assembly implementation of
efi_call(), and only keep the pieces that deal with the stack
alignment and the ABI translation.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Arvind Sankar <nivedita@alum.mit.edu>
Cc: Matthew Garrett <mjg59@google.com>
Cc: linux-efi@vger.kernel.org
Link: https://lkml.kernel.org/r/20200103113953.9571-10-ardb@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The variadic efi_call_phys() wrapper that exists on i386 was
originally created to call into any EFI firmware runtime service,
but in practice, we only use it once, to call SetVirtualAddressMap()
during early boot.
The flexibility provided by the variadic nature also makes it
type unsafe, and makes the assembler code more complicated than
needed, since it has to deal with an unknown number of arguments
living on the stack.
So clean this up, by renaming the helper to efi_call_svam(), and
dropping the unneeded complexity. Let's also drop the reference
to the efi_phys struct and grab the address from the EFI system
table directly.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Arvind Sankar <nivedita@alum.mit.edu>
Cc: Matthew Garrett <mjg59@google.com>
Cc: linux-efi@vger.kernel.org
Link: https://lkml.kernel.org/r/20200103113953.9571-9-ardb@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Split the phys_efi_set_virtual_address_map() routine into 32 and 64 bit
versions, so we can simplify them individually in subsequent patches.
There is very little overlap between the logic anyway, and this has
already been factored out in prolog/epilog routines which are completely
different between 32 bit and 64 bit. So let's take it one step further,
and get rid of the overlap completely.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Arvind Sankar <nivedita@alum.mit.edu>
Cc: Matthew Garrett <mjg59@google.com>
Cc: linux-efi@vger.kernel.org
Link: https://lkml.kernel.org/r/20200103113953.9571-8-ardb@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
In a subsequent patch, we will fold the prolog/epilog routines that are
part of the support code to call SetVirtualAddressMap() with a 1:1
mapping into the callers. However, the 64-bit version mostly consists
of ugly mapping code that is only used when efi=old_map is in effect,
which is extremely rare. So let's move this code out of the way so it
does not clutter the common code.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Arvind Sankar <nivedita@alum.mit.edu>
Cc: Matthew Garrett <mjg59@google.com>
Cc: linux-efi@vger.kernel.org
Link: https://lkml.kernel.org/r/20200103113953.9571-7-ardb@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
All EFI firmware call prototypes have been annotated as __efiapi,
permitting us to attach attributes regarding the calling convention
by overriding __efiapi to an architecture specific value.
On 32-bit x86, EFI firmware calls use the plain calling convention
where all arguments are passed via the stack, and cleaned up by the
caller. Let's add this to the __efiapi definition so we no longer
need to cast the function pointers before invoking them.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Arvind Sankar <nivedita@alum.mit.edu>
Cc: Matthew Garrett <mjg59@google.com>
Cc: linux-efi@vger.kernel.org
Link: https://lkml.kernel.org/r/20200103113953.9571-6-ardb@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Fix a couple of issues with the way we map and copy the vendor string:
- we map only 2 bytes, which usually works since you get at least a
page, but if the vendor string happens to cross a page boundary,
a crash will result
- only call early_memunmap() if early_memremap() succeeded, or we will
call it with a NULL address which it doesn't like,
- while at it, switch to early_memremap_ro(), and array indexing rather
than pointer dereferencing to read the CHAR16 characters.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Arvind Sankar <nivedita@alum.mit.edu>
Cc: Matthew Garrett <mjg59@google.com>
Cc: linux-efi@vger.kernel.org
Fixes: 5b83683f32 ("x86: EFI runtime service support")
Link: https://lkml.kernel.org/r/20200103113953.9571-5-ardb@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Commit a8147dba75 ("efi/x86: Rename efi_is_native() to efi_is_mixed()")
renamed and refactored efi_is_native() into efi_is_mixed(), but failed
to take into account that these are not diametrical opposites.
Mixed mode is a construct that permits 64-bit kernels to boot on 32-bit
firmware, but there is another non-native combination which is supported,
i.e., 32-bit kernels booting on 64-bit firmware, but only for boot and not
for runtime services. Also, mixed mode can be disabled in Kconfig, in
which case the 64-bit kernel can still be booted from 32-bit firmware,
but without access to runtime services.
Due to this oversight, efi_runtime_supported() now incorrectly returns
true for such configurations, resulting in crashes at boot. So fix this
by making efi_runtime_supported() aware of this.
As a side effect, some efi_thunk_xxx() stubs have become obsolete, so
remove them as well.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Arvind Sankar <nivedita@alum.mit.edu>
Cc: Matthew Garrett <mjg59@google.com>
Cc: linux-efi@vger.kernel.org
Link: https://lkml.kernel.org/r/20200103113953.9571-4-ardb@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Commit c3710de506 ("efi/libstub/x86: Drop __efi_early() export and
efi_config struct") introduced a reference from C code in eboot.c to
the startup_32 symbol defined in the .S startup code. This results in
a GOT based reference to startup_32, and since GOT entries carry
absolute addresses, they need to be fixed up before they can be used.
On modern toolchains (binutils 2.26 or later), this reference is
relaxed into a R_386_GOTOFF relocation (or the analogous X86_64 one)
which never uses the absolute address in the entry, and so we get
away with not fixing up the GOT table before calling the EFI entry
point. However, GCC 4.6 combined with a binutils of the era (2.24)
will produce a true GOT indirected reference, resulting in a wrong
value to be returned for the address of startup_32() if the boot
code is not running at the address it was linked at.
Fortunately, we can easily override this behavior, and force GCC to
emit the GOTOFF relocations explicitly, by setting the visibility
pragma 'hidden'.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Arvind Sankar <nivedita@alum.mit.edu>
Cc: Matthew Garrett <mjg59@google.com>
Cc: linux-efi@vger.kernel.org
Link: https://lkml.kernel.org/r/20200103113953.9571-3-ardb@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The mixed mode refactor actually broke mixed mode by failing to
pass the bootparam structure to startup_32(). This went unnoticed
because it apparently has a high tolerance for being passed random
junk, and still boots fine in some cases. So let's fix this by
populating %esi as required when entering via efi32_stub_entry,
and while at it, preserve the arguments themselves instead of their
address in memory (via the stack pointer) since that memory could
be clobbered before we get to it.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Arvind Sankar <nivedita@alum.mit.edu>
Cc: Matthew Garrett <mjg59@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-efi@vger.kernel.org
Link: https://lkml.kernel.org/r/20200103113953.9571-2-ardb@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Use devm_platform_ioremap_resource() to simplify the code a bit.
While here, drop initialized but unused ssram_base_addr and ssram_size members.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
The ungrafting from PRIO bug fixes in net, when merged into net-next,
merge cleanly but create a build failure. The resolution used here is
from Petr Machata.
Signed-off-by: David S. Miller <davem@davemloft.net>
The patch introduces BPF_MAP_TYPE_STRUCT_OPS. The map value
is a kernel struct with its func ptr implemented in bpf prog.
This new map is the interface to register/unregister/introspect
a bpf implemented kernel struct.
The kernel struct is actually embedded inside another new struct
(or called the "value" struct in the code). For example,
"struct tcp_congestion_ops" is embbeded in:
struct bpf_struct_ops_tcp_congestion_ops {
refcount_t refcnt;
enum bpf_struct_ops_state state;
struct tcp_congestion_ops data; /* <-- kernel subsystem struct here */
}
The map value is "struct bpf_struct_ops_tcp_congestion_ops".
The "bpftool map dump" will then be able to show the
state ("inuse"/"tobefree") and the number of subsystem's refcnt (e.g.
number of tcp_sock in the tcp_congestion_ops case). This "value" struct
is created automatically by a macro. Having a separate "value" struct
will also make extending "struct bpf_struct_ops_XYZ" easier (e.g. adding
"void (*init)(void)" to "struct bpf_struct_ops_XYZ" to do some
initialization works before registering the struct_ops to the kernel
subsystem). The libbpf will take care of finding and populating the
"struct bpf_struct_ops_XYZ" from "struct XYZ".
Register a struct_ops to a kernel subsystem:
1. Load all needed BPF_PROG_TYPE_STRUCT_OPS prog(s)
2. Create a BPF_MAP_TYPE_STRUCT_OPS with attr->btf_vmlinux_value_type_id
set to the btf id "struct bpf_struct_ops_tcp_congestion_ops" of the
running kernel.
Instead of reusing the attr->btf_value_type_id,
btf_vmlinux_value_type_id s added such that attr->btf_fd can still be
used as the "user" btf which could store other useful sysadmin/debug
info that may be introduced in the furture,
e.g. creation-date/compiler-details/map-creator...etc.
3. Create a "struct bpf_struct_ops_tcp_congestion_ops" object as described
in the running kernel btf. Populate the value of this object.
The function ptr should be populated with the prog fds.
4. Call BPF_MAP_UPDATE with the object created in (3) as
the map value. The key is always "0".
During BPF_MAP_UPDATE, the code that saves the kernel-func-ptr's
args as an array of u64 is generated. BPF_MAP_UPDATE also allows
the specific struct_ops to do some final checks in "st_ops->init_member()"
(e.g. ensure all mandatory func ptrs are implemented).
If everything looks good, it will register this kernel struct
to the kernel subsystem. The map will not allow further update
from this point.
Unregister a struct_ops from the kernel subsystem:
BPF_MAP_DELETE with key "0".
Introspect a struct_ops:
BPF_MAP_LOOKUP_ELEM with key "0". The map value returned will
have the prog _id_ populated as the func ptr.
The map value state (enum bpf_struct_ops_state) will transit from:
INIT (map created) =>
INUSE (map updated, i.e. reg) =>
TOBEFREE (map value deleted, i.e. unreg)
The kernel subsystem needs to call bpf_struct_ops_get() and
bpf_struct_ops_put() to manage the "refcnt" in the
"struct bpf_struct_ops_XYZ". This patch uses a separate refcnt
for the purose of tracking the subsystem usage. Another approach
is to reuse the map->refcnt and then "show" (i.e. during map_lookup)
the subsystem's usage by doing map->refcnt - map->usercnt to filter out
the map-fd/pinned-map usage. However, that will also tie down the
future semantics of map->refcnt and map->usercnt.
The very first subsystem's refcnt (during reg()) holds one
count to map->refcnt. When the very last subsystem's refcnt
is gone, it will also release the map->refcnt. All bpf_prog will be
freed when the map->refcnt reaches 0 (i.e. during map_free()).
Here is how the bpftool map command will look like:
[root@arch-fb-vm1 bpf]# bpftool map show
6: struct_ops name dctcp flags 0x0
key 4B value 256B max_entries 1 memlock 4096B
btf_id 6
[root@arch-fb-vm1 bpf]# bpftool map dump id 6
[{
"value": {
"refcnt": {
"refs": {
"counter": 1
}
},
"state": 1,
"data": {
"list": {
"next": 0,
"prev": 0
},
"key": 0,
"flags": 2,
"init": 24,
"release": 0,
"ssthresh": 25,
"cong_avoid": 30,
"set_state": 27,
"cwnd_event": 28,
"in_ack_event": 26,
"undo_cwnd": 29,
"pkts_acked": 0,
"min_tso_segs": 0,
"sndbuf_expand": 0,
"cong_control": 0,
"get_info": 0,
"name": [98,112,102,95,100,99,116,99,112,0,0,0,0,0,0,0
],
"owner": 0
}
}
}
]
Misc Notes:
* bpf_struct_ops_map_sys_lookup_elem() is added for syscall lookup.
It does an inplace update on "*value" instead returning a pointer
to syscall.c. Otherwise, it needs a separate copy of "zero" value
for the BPF_STRUCT_OPS_STATE_INIT to avoid races.
* The bpf_struct_ops_map_delete_elem() is also called without
preempt_disable() from map_delete_elem(). It is because
the "->unreg()" may requires sleepable context, e.g.
the "tcp_unregister_congestion_control()".
* "const" is added to some of the existing "struct btf_func_model *"
function arg to avoid a compiler warning caused by this patch.
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Andrii Nakryiko <andriin@fb.com>
Acked-by: Yonghong Song <yhs@fb.com>
Link: https://lore.kernel.org/bpf/20200109003505.3855919-1-kafai@fb.com
Use resource_size() rather than a verbose computation on
the end and start fields.
The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)
<smpl>
@@ struct resource ptr; @@
- (ptr.end - ptr.start + 1)
+ resource_size(&ptr)
</smpl>
Signed-off-by: Julia Lawall <Julia.Lawall@inria.fr>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/1577900990-8588-10-git-send-email-Julia.Lawall@inria.fr
ignore_sysret() contains an unsuffixed SYSRET instruction. gas correctly
interprets this as SYSRETL, but leaving it up to gas to guess when there
is no register operand that implies a size is bad practice, and upstream
gas is likely to warn about this in the future. Use SYSRETL explicitly.
This does not change the assembled output.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Acked-by: Andy Lutomirski <luto@kernel.org>
Link: https://lkml.kernel.org/r/038a7c35-062b-a285-c6d2-653b56585844@suse.com
The CRYPTO_TFM_RES_* flags were apparently meant as a way to make the
->setkey() functions provide more information about errors. But these
flags weren't actually being used or tested, and in many cases they
weren't being set correctly anyway. So they've now been removed.
Also, if someone ever actually needs to start better distinguishing
->setkey() errors (which is somewhat unlikely, as this has been unneeded
for a long time), we'd be much better off just defining different return
values, like -EINVAL if the key is invalid for the algorithm vs.
-EKEYREJECTED if the key was rejected by a policy like "no weak keys".
That would be much simpler, less error-prone, and easier to test.
So just remove CRYPTO_TFM_RES_MASK and all the unneeded logic that
propagates these flags around.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The CRYPTO_TFM_RES_BAD_KEY_LEN flag was apparently meant as a way to
make the ->setkey() functions provide more information about errors.
However, no one actually checks for this flag, which makes it pointless.
Also, many algorithms fail to set this flag when given a bad length key.
Reviewing just the generic implementations, this is the case for
aes-fixed-time, cbcmac, echainiv, nhpoly1305, pcrypt, rfc3686, rfc4309,
rfc7539, rfc7539esp, salsa20, seqiv, and xcbc. But there are probably
many more in arch/*/crypto/ and drivers/crypto/.
Some algorithms can even set this flag when the key is the correct
length. For example, authenc and authencesn set it when the key payload
is malformed in any way (not just a bad length), the atmel-sha and ccree
drivers can set it if a memory allocation fails, and the chelsio driver
sets it for bad auth tag lengths, not just bad key lengths.
So even if someone actually wanted to start checking this flag (which
seems unlikely, since it's been unused for a long time), there would be
a lot of work needed to get it working correctly. But it would probably
be much better to go back to the drawing board and just define different
return values, like -EINVAL if the key is invalid for the algorithm vs.
-EKEYREJECTED if the key was rejected by a policy like "no weak keys".
That would be much simpler, less error-prone, and easier to test.
So just remove this flag.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Reviewed-by: Horia Geantă <horia.geanta@nxp.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
force_iret() was originally intended to prevent the return to user mode with
the SYSRET or SYSEXIT instructions, in cases where the register state could
have been changed to be incompatible with those instructions. The entry code
has been significantly reworked since then, and register state is validated
before SYSRET or SYSEXIT are used. force_iret() no longer serves its original
purpose and can be eliminated.
Signed-off-by: Brian Gerst <brgerst@gmail.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Acked-by: Oleg Nesterov <oleg@redhat.com>
Link: https://lkml.kernel.org/r/20191219115812.102620-1-brgerst@gmail.com
>From the Intel Optimization Reference Manual:
3.7.6.1 Fast Short REP MOVSB
Beginning with processors based on Ice Lake Client microarchitecture,
REP MOVSB performance of short operations is enhanced. The enhancement
applies to string lengths between 1 and 128 bytes long. Support for
fast-short REP MOVSB is enumerated by the CPUID feature flag: CPUID
[EAX=7H, ECX=0H).EDX.FAST_SHORT_REP_MOVSB[bit 4] = 1. There is no change
in the REP STOS performance.
Add an X86_FEATURE_FSRM flag for this.
memmove() avoids REP MOVSB for short (< 32 byte) copies. Check FSRM and
use REP MOVSB for short copies on systems that support it.
[ bp: Massage and add comment. ]
Signed-off-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20191216214254.26492-1-tony.luck@intel.com
Chen Yu