search for: mul_u64_u64_shr

...t_tick, cur_tsc; > + > + while (1) { > + sequence = READ_ONCE(tsc_pg->tsc_sequence); > + if (!sequence) > + break; > + > + scale = READ_ONCE(tsc_pg->tsc_scale); > + offset = READ_ONCE(tsc_pg->tsc_offset); > + rdtscll(cur_tsc); > + > + current_tick = mul_u64_u64_shr(cur_tsc, scale, 64) + offset; > + > + if (READ_ONCE(tsc_pg->tsc_sequence) == sequence) > + return current_tick; That sequence stuff lacks still a sensible explanation. It's fundamentally different from the sequence counting we do in the kernel, so documentation for it is really...

...t_tick, cur_tsc; > + > + while (1) { > + sequence = READ_ONCE(tsc_pg->tsc_sequence); > + if (!sequence) > + break; > + > + scale = READ_ONCE(tsc_pg->tsc_scale); > + offset = READ_ONCE(tsc_pg->tsc_offset); > + rdtscll(cur_tsc); > + > + current_tick = mul_u64_u64_shr(cur_tsc, scale, 64) + offset; > + > + if (READ_ONCE(tsc_pg->tsc_sequence) == sequence) > + return current_tick; That sequence stuff lacks still a sensible explanation. It's fundamentally different from the sequence counting we do in the kernel, so documentation for it is really...

...> + scale = READ_ONCE(tsc_pg->tsc_scale); > + offset = READ_ONCE(tsc_pg->tsc_offset); > + cur_tsc = rdtsc_ordered(); Since you already have smp_ barriers and rdtsc_ordered is a barrier, the compiler barriers (READ_ONCE()) shouldn't be necessary. > + > + current_tick = mul_u64_u64_shr(cur_tsc, scale, 64) + offset; > + > + /* > + * Make sure we read sequence after we read all other values > + * from TSC page. > + */ > + smp_rmb(); > + > + if (READ_ONCE(tsc_pg->tsc_sequence) == sequence) > + return current_tick; > + } Why not make do {...

...> + scale = READ_ONCE(tsc_pg->tsc_scale); > + offset = READ_ONCE(tsc_pg->tsc_offset); > + cur_tsc = rdtsc_ordered(); Since you already have smp_ barriers and rdtsc_ordered is a barrier, the compiler barriers (READ_ONCE()) shouldn't be necessary. > + > + current_tick = mul_u64_u64_shr(cur_tsc, scale, 64) + offset; > + > + /* > + * Make sure we read sequence after we read all other values > + * from TSC page. > + */ > + smp_rmb(); > + > + if (READ_ONCE(tsc_pg->tsc_sequence) == sequence) > + return current_tick; > + } Why not make do {...

...ence) >> > + break; >> > + >> > + scale = READ_ONCE(tsc_pg->tsc_scale); >> > + offset = READ_ONCE(tsc_pg->tsc_offset); >> > + rdtscll(cur_tsc); >> > + >> > + current_tick = mul_u64_u64_shr(cur_tsc, scale, 64) + offset; >> > + >> > + if (READ_ONCE(tsc_pg->tsc_sequence) == sequence) >> > + return current_tick; >> >> That sequence stuff lacks still a sensible explanation. It's fundamentally >> different from...

...ence) >> > + break; >> > + >> > + scale = READ_ONCE(tsc_pg->tsc_scale); >> > + offset = READ_ONCE(tsc_pg->tsc_offset); >> > + rdtscll(cur_tsc); >> > + >> > + current_tick = mul_u64_u64_shr(cur_tsc, scale, 64) + offset; >> > + >> > + if (READ_ONCE(tsc_pg->tsc_sequence) == sequence) >> > + return current_tick; >> >> That sequence stuff lacks still a sensible explanation. It's fundamentally >> different from...

...h it. As nobody seems to be strongly offended by my RFC I'm sending first non-RFC version. Patch 1 is made on top of K. Y.'s code refactoring which moved tsc page clocksource to arch/x86/hyperv/hv_init.c, this is currently present in Greg's char-misc-next tree. Changes since RFC: - Use mul_u64_u64_shr() instead of an open coded implementation [Andy Lutomirski, Thomas Gleixner] - Don't use the same pvclock_page for both VCLOCK_PVCLOCK and VCLOCK_HVCLOCK, create another one [Andy Lutomirski] - Fix issues reported by kbuild test robot. - Rename HYPERV_CLOCK -> HYPERV_TSCPAGE to avoid the...

...h it. As nobody seems to be strongly offended by my RFC I'm sending first non-RFC version. Patch 1 is made on top of K. Y.'s code refactoring which moved tsc page clocksource to arch/x86/hyperv/hv_init.c, this is currently present in Greg's char-misc-next tree. Changes since RFC: - Use mul_u64_u64_shr() instead of an open coded implementation [Andy Lutomirski, Thomas Gleixner] - Don't use the same pvclock_page for both VCLOCK_PVCLOCK and VCLOCK_HVCLOCK, create another one [Andy Lutomirski] - Fix issues reported by kbuild test robot. - Rename HYPERV_CLOCK -> HYPERV_TSCPAGE to avoid the...

...page reading logic to asm/mshyperv.h. While on it, do the following - Document the reading algorithm. - Simplify the code a bit. - Add explicit READ_ONCE() to not rely on 'volatile'. - Add explicit barriers to prevent re-ordering (we need to read sequence strictly before and after) - Use mul_u64_u64_shr() instead of assembly, gcc generates a single 'mul' instruction on x86_64 anyway. Signed-off-by: Vitaly Kuznetsov <vkuznets at redhat.com> --- arch/x86/hyperv/hv_init.c | 36 ++++------------------------- arch/x86/include/asm/mshyperv.h | 50 +++++++++++++++++++++++++++++++++...

Hi, merge window is about to close so I hope it's OK to make another try here. Changes since v2: - Add explicit READ_ONCE() to not rely on 'volatile' [Andy Lutomirski] - rdtsc() -> rdtsc_ordered() [Andy Lutomirski] - virt_rmb() -> smp_rmb() [Thomas Gleixner, Andy Lutomirski] Thomas, Andy, it seems the only blocker for the series was the ambiguity with TSC page read algorithm.

Hi, merge window is about to close so I hope it's OK to make another try here. Changes since v2: - Add explicit READ_ONCE() to not rely on 'volatile' [Andy Lutomirski] - rdtsc() -> rdtsc_ordered() [Andy Lutomirski] - virt_rmb() -> smp_rmb() [Thomas Gleixner, Andy Lutomirski] Thomas, Andy, it seems the only blocker for the series was the ambiguity with TSC page read algorithm.

...t; > + sequence = READ_ONCE(tsc_pg->tsc_sequence); > > + if (!sequence) > > + break; > > + > > + scale = READ_ONCE(tsc_pg->tsc_scale); > > + offset = READ_ONCE(tsc_pg->tsc_offset); > > + rdtscll(cur_tsc); > > + > > + current_tick = mul_u64_u64_shr(cur_tsc, scale, 64) + offset; > > + > > + if (READ_ONCE(tsc_pg->tsc_sequence) == sequence) > > + return current_tick; > > That sequence stuff lacks still a sensible explanation. It's fundamentally > different from the sequence counting we do in the kernel, so d...

...(1) { >> + sequence = READ_ONCE(tsc_pg->tsc_sequence); >> + if (!sequence) >> + break; >> + >> + scale = READ_ONCE(tsc_pg->tsc_scale); >> + offset = READ_ONCE(tsc_pg->tsc_offset); >> + rdtscll(cur_tsc); >> + >> + current_tick = mul_u64_u64_shr(cur_tsc, scale, 64) + offset; >> + >> + if (READ_ONCE(tsc_pg->tsc_sequence) == sequence) >> + return current_tick; > > That sequence stuff lacks still a sensible explanation. It's fundamentally > different from the sequence counting we do in the kernel, so docum...

...t_tick, cur_tsc; > + > + while (1) { > + sequence = READ_ONCE(tsc_pg->tsc_sequence); > + if (!sequence) > + break; > + > + scale = READ_ONCE(tsc_pg->tsc_scale); > + offset = READ_ONCE(tsc_pg->tsc_offset); > + rdtscll(cur_tsc); > + > + current_tick = mul_u64_u64_shr(cur_tsc, scale, 64) + offset; > + > + if (READ_ONCE(tsc_pg->tsc_sequence) == sequence) > + return current_tick; That sequence stuff lacks still a sensible explanation. It's fundamentally different from the sequence counting we do in the kernel, so documentation for it is really...

...amp;hvclock_page; + u64 sequence, scale, offset, current_tick, cur_tsc; + + while (1) { + sequence = READ_ONCE(tsc_pg->tsc_sequence); + if (!sequence) + break; + + scale = READ_ONCE(tsc_pg->tsc_scale); + offset = READ_ONCE(tsc_pg->tsc_offset); + rdtscll(cur_tsc); + + current_tick = mul_u64_u64_shr(cur_tsc, scale, 64) + offset; + + if (READ_ONCE(tsc_pg->tsc_sequence) == sequence) + return current_tick; + } + + *mode = VCLOCK_NONE; + return 0; +} +#endif notrace static u64 vread_tsc(void) { @@ -173,6 +205,10 @@ notrace static inline u64 vgetsns(int *mode) else if (gtod->vclock_m...

Hi, while we're still waiting for a definitive ACK from Microsoft that the algorithm is good for SMP case (as we can't prevent the code in vdso from migrating between CPUs) I'd like to send v2 with some modifications to keep the discussion going. Changes since v1: - Document the TSC page reading protocol [Thomas Gleixner]. - Separate the TSC page reading code from

Hi, while we're still waiting for a definitive ACK from Microsoft that the algorithm is good for SMP case (as we can't prevent the code in vdso from migrating between CPUs) I'd like to send v2 with some modifications to keep the discussion going. Changes since v1: - Document the TSC page reading protocol [Thomas Gleixner]. - Separate the TSC page reading code from

On Wed, Feb 8, 2017 at 9:07 AM, Vitaly Kuznetsov <vkuznets at redhat.com> wrote: > Hyper-V TSC page clocksource is suitable for vDSO, however, the protocol > defined by the hypervisor is different from VCLOCK_PVCLOCK. Implement the > required support re-using pvclock_page VVAR as VCLOCK_PVCLOCK is mutually > exclusive with VCLOCK_HVCLOCK at run time. > > Signed-off-by:

On Wed, Feb 8, 2017 at 9:07 AM, Vitaly Kuznetsov <vkuznets at redhat.com> wrote: > Hyper-V TSC page clocksource is suitable for vDSO, however, the protocol > defined by the hypervisor is different from VCLOCK_PVCLOCK. Implement the > required support re-using pvclock_page VVAR as VCLOCK_PVCLOCK is mutually > exclusive with VCLOCK_HVCLOCK at run time. > > Signed-off-by: