MIPS already uses and S390 will need the vdso data pointer in __arch_get_hw_counter(). This works nicely as long as the architecture does not support time namespaces in the VDSO. With time namespaces enabled the regular accessor to the vdso data pointer __arch_get_vdso_data() will return the namespace specific VDSO data page for tasks which are part of a non-root time namespace. This would cause the architectures which need the vdso data pointer in __arch_get_hw_counter() to access the wrong vdso data page. Add a vdso_data pointer argument to __arch_get_hw_counter() and hand it in from the call sites in the core code. For architectures which do not need the data pointer in their counter accessor function the compiler will just optimize it out. Fix up all existing architecture implementations and make MIPS utilize the pointer instead of invoking the accessor function. No functional change and no change in the resulting object code (except MIPS). Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Link: https://lkml.kernel.org/r/draft-87wo2ekuzn.fsf@nanos.tec.linutronix.de
Pull more x86 updates from Thomas Gleixner:
"A set of fixes and updates for x86:
- Unbreak paravirt VDSO clocks.
While the VDSO code was moved into lib for sharing a subtle check
for the validity of paravirt clocks got replaced. While the
replacement works perfectly fine for bare metal as the update of
the VDSO clock mode is synchronous, it fails for paravirt clocks
because the hypervisor can invalidate them asynchronously.
Bring it back as an optional function so it does not inflict this
on architectures which are free of PV damage.
- Fix the jiffies to jiffies64 mapping on 64bit so it does not
trigger an ODR violation on newer compilers
- Three fixes for the SSBD and *IB* speculation mitigation maze to
ensure consistency, not disabling of some *IB* variants wrongly and
to prevent a rogue cross process shutdown of SSBD. All marked for
stable.
- Add yet more CPU models to the splitlock detection capable list
!@#%$!
- Bring the pr_info() back which tells that TSC deadline timer is
enabled.
- Reboot quirk for MacBook6,1"
* tag 'x86-urgent-2020-06-11' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/vdso: Unbreak paravirt VDSO clocks
lib/vdso: Provide sanity check for cycles (again)
clocksource: Remove obsolete ifdef
x86_64: Fix jiffies ODR violation
x86/speculation: PR_SPEC_FORCE_DISABLE enforcement for indirect branches.
x86/speculation: Prevent rogue cross-process SSBD shutdown
x86/speculation: Avoid force-disabling IBPB based on STIBP and enhanced IBRS.
x86/cpu: Add Sapphire Rapids CPU model number
x86/split_lock: Add Icelake microserver and Tigerlake CPU models
x86/apic: Make TSC deadline timer detection message visible
x86/reboot/quirks: Add MacBook6,1 reboot quirk
The original x86 VDSO implementation checked for the validity of the clock
source read by testing whether the returned signed cycles value is less
than zero. This check was also used by the vdso read function to signal
that the current selected clocksource is not VDSO capable.
During the rework of the VDSO code the check was removed and replaced with
a check for the clocksource mode being != NONE.
This turned out to be a mistake because the check is necessary for paravirt
and hyperv clock sources. The reason is that these clock sources have their
own internal sequence counter to validate the clocksource at the point of
reading it. This is necessary because the hypervisor can invalidate the
clocksource asynchronously so a check during the VDSO data update is not
sufficient. Having a separate indicator for the validity is slower than
just validating the cycles value. The check for it being negative turned
out to be the fastest implementation and safe as it would require an uptime
of ~73 years with a 4GHz counter frequency to result in a false positive.
Add an optional function to validate the cycles with a default
implementation which allows the compiler to optimize it out for
architectures which do not require it.
Fixes: 5d51bee725cc ("clocksource: Add common vdso clock mode storage")
Reported-by: Miklos Szeredi <miklos@szeredi.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Miklos Szeredi <mszeredi@redhat.com>
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20200606221531.963970768@linutronix.de
When adding gettime64() to a 32 bit architecture (namely powerpc/32) it has been noticed that GCC doesn't inline anymore __cvdso_clock_gettime_common() because it is called twice (Once by __cvdso_clock_gettime() and once by __cvdso_clock_gettime32). This has the effect of seriously degrading the performance: Before the implementation of gettime64(), gettime() runs in: clock-gettime-monotonic-raw: vdso: 1003 nsec/call clock-gettime-monotonic-coarse: vdso: 592 nsec/call clock-gettime-monotonic: vdso: 942 nsec/call When adding a gettime64() entry point, the standard gettime() performance is degraded by 30% to 50%: clock-gettime-monotonic-raw: vdso: 1300 nsec/call clock-gettime-monotonic-coarse: vdso: 900 nsec/call clock-gettime-monotonic: vdso: 1232 nsec/call Adding __always_inline() to __cvdso_clock_gettime_common() regains the original performance. In terms of code size, the inlining increases the code size by only 176 bytes. This is in the noise for a kernel image. Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Link: https://lkml.kernel.org/r/1ab6a62c356c3bec35d1623563ef9c636205bcda.1588079622.git.christophe.leroy@c-s.fr
The vDSO library should only include the necessary headers required for a userspace library (UAPI and a minimal set of kernel headers). To make this possible it is necessary to isolate from the kernel headers the common parts that are strictly necessary to build the library. Refactor the unified vdso code to use the common headers. Signed-off-by: Vincenzo Frascino <vincenzo.frascino@arm.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Link: https://lkml.kernel.org/r/20200320145351.32292-26-vincenzo.frascino@arm.com
On powerpc, __arch_get_vdso_data() clobbers the link register, requiring the caller to save it. As the parent function already has to set a stack frame and saves the link register before calling the C vdso function, retrieving the vdso data pointer there is less overhead. Split out the functional code from the __cvdso.*() interfaces into new static functions which can either be called from the existing interfaces with the vdso data pointer supplied via __arch_get_vdso_data() or directly from ASM code. Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Tested-by: Vincenzo Frascino <vincenzo.frascino@arm.com> Reviewed-by: Vincenzo Frascino <vincenzo.frascino@arm.com> Link: https://lore.kernel.org/r/abf97996602ef07223fec30c005df78e5ed41b2e.1580399657.git.christophe.leroy@c-s.fr Link: https://lkml.kernel.org/r/20200207124403.965789141@linutronix.de
On powerpc/32, GCC (8.1) generates pretty bad code for the ns >>= vd->shift
operation taking into account that the shift is always <= 32 and the upper
part of the result is likely to be zero. GCC makes reversed assumptions
considering the shift to be likely >= 32 and the upper part to be like not
zero.
unsigned long long shift(unsigned long long x, unsigned char s)
{
return x >> s;
}
results in:
00000018 <shift>:
18: 35 25 ff e0 addic. r9,r5,-32
1c: 41 80 00 10 blt 2c <shift+0x14>
20: 7c 64 4c 30 srw r4,r3,r9
24: 38 60 00 00 li r3,0
28: 4e 80 00 20 blr
2c: 54 69 08 3c rlwinm r9,r3,1,0,30
30: 21 45 00 1f subfic r10,r5,31
34: 7c 84 2c 30 srw r4,r4,r5
38: 7d 29 50 30 slw r9,r9,r10
3c: 7c 63 2c 30 srw r3,r3,r5
40: 7d 24 23 78 or r4,r9,r4
44: 4e 80 00 20 blr
Even when forcing the shift to be smaller than 32 with an &= 31, it still
considers the shift as likely >= 32.
Move the default shift implementation into an inline which can be redefined
in architecture code via a macro.
[ tglx: Made the shift argument u32 and removed the __arch prefix ]
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Vincenzo Frascino <vincenzo.frascino@arm.com>
Reviewed-by: Vincenzo Frascino <vincenzo.frascino@arm.com>
Link: https://lore.kernel.org/r/b3d449de856982ed060a71e6ace8eeca4654e685.1580399657.git.christophe.leroy@c-s.fr
Link: https://lkml.kernel.org/r/20200207124403.857649978@linutronix.de
Some architectures have a fixed clocksource which is known at compile time and cannot be replaced or disabled at runtime, e.g. timebase on PowerPC. For such cases the clock mode check in the VDSO code is pointless. Move the check for a VDSO capable clocksource into an inline function and allow architectures to redefine it via a macro. [ tglx: Removed the #ifdef mess ] Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Tested-by: Vincenzo Frascino <vincenzo.frascino@arm.com> Reviewed-by: Vincenzo Frascino <vincenzo.frascino@arm.com> Link: https://lkml.kernel.org/r/20200207124403.748756829@linutronix.de
Move the time namespace indicator clock mode to the other ones for consistency sake. Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Reviewed-by: Vincenzo Frascino <vincenzo.frascino@arm.com> Link: https://lkml.kernel.org/r/20200207124403.656097274@linutronix.de
Now that all architectures are converted to use the generic storage the helpers and conditionals can be removed. Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Tested-by: Vincenzo Frascino <vincenzo.frascino@arm.com> Reviewed-by: Vincenzo Frascino <vincenzo.frascino@arm.com> Link: https://lkml.kernel.org/r/20200207124403.470699892@linutronix.de