x86/hyperv: Replace kmap() with kmap_local_page()
Subsystems (3)
PATCH 1/1 x86/hyperv: Replace kmap() with kmap_local_page()
From: Zhao Liu <[email protected]>
kmap() is being deprecated in favor of kmap_local_page()[1].
There are two main problems with kmap(): (1) It comes with an overhead as mapping space is restricted and protected by a global lock for synchronization and (2) it also requires global TLB invalidation when the kmap's pool wraps and it might block when the mapping space is fully utilized until a slot becomes available.
With kmap_local_page() the mappings are per thread, CPU local, can take page faults, and can be called from any context (including interrupts). It is faster than kmap() in kernels with HIGHMEM enabled. Furthermore, the tasks can be preempted and, when they are scheduled to run again, the kernel virtual addresses are restored and are still valid.
In the fuction hyperv_init() of hyperv/hv_init.c, the mapping is used in a single thread and is short live. So, in this case, it's safe to simply use kmap_local_page() to create mapping, and this avoids the wasted cost of kmap() for global synchronization.
In addtion, the fuction hyperv_init() checks if kmap() fails by BUG_ON(). From the original discussion[2], the BUG_ON() here is just used to explicitly panic NULL pointer. So still keep the BUG_ON() in place to check if kmap_local_page() fails. Based on this consideration, memcpy_to_page() is not selected here but only kmap_local_page() is used.
Therefore, replace kmap() with kmap_local_page() in hyperv/hv_init.c.
[1]: https://lore.kernel.org/all/[email protected]
[2]: https://lore.kernel.org/lkml/20200915103710.cqmdvzh5lys4wsqo@liuwe-devbox-debian-v2/
---
Suggested by credits.
Dave: Referred to his comments about whether kmap() can fail and the
suggestion to keep BUG_ON() in place.
Ira: Referred to his task documentation and review comments about
keeping BUG_ON() for kmap_local_page().
Fabio: Stole some of his boiler plate commit message.
--- arch/x86/hyperv/hv_init.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
arch/x86/hyperv/hv_init.c +2 -2
@@ -459,13 +459,13 @@ void __init hyperv_init(void)wrmsrl (HV_X64_MSR_HYPERCALL , hypercall_msr.as_uint64); pg =vmalloc_to_page (hv_hypercall_pg); - dst =kmap (pg); + dst =kmap_local_page (pg); src =memremap (hypercall_msr.guest_physical_address <<PAGE_SHIFT ,PAGE_SIZE ,MEMREMAP_WB );BUG_ON (!(src && dst));memcpy (dst, src,HV_HYP_PAGE_SIZE );memunmap (src); -kunmap (pg); +kunmap_local (dst); }else { hypercall_msr.guest_physical_address =vmalloc_to_pfn (hv_hypercall_pg);wrmsrl (HV_X64_MSR_HYPERCALL , hypercall_msr.as_uint64);
Signature · 1 line
2.34.1
Wei Liu
maintainerQuoted 16 lines · Zhao Liu
From: Zhao Liu <[email protected]>
kmap()is being deprecated in favor of kmap_local_page()[1].There are two main problems with
kmap(): (1) It comes with an overhead as mapping space is restricted and protected by a global lock for synchronization and (2) it also requires global TLB invalidation when the kmap's pool wraps and it might block when the mapping space is fully utilized until a slot becomes available.With
kmap_local_page()the mappings are per thread, CPU local, can take page faults, and can be called from any context (including interrupts). It is faster thankmap()in kernels with HIGHMEM enabled. Furthermore, the tasks can be preempted and, when they are scheduled to run again, the kernel virtual addresses are restored and are still valid.
Quoted 5 lines
In the fuction
hyperv_init()ofhyperv/hv_init.c, the mapping is used in a single thread and is short live. So, in this case, it's safe to simply usekmap_local_page()to create mapping, and this avoids the wasted cost ofkmap()for global synchronization.
The kmap call in that function is not performance critical in any way, and at this point in the initialization process I don't expect there to be any contention, so the downside of kmap is not really a concern here.
That being said, kmap getting deprecated is a good enough reason to switch to kmap_local_page. And I appreciate this well-written, well-reasoned commit message.
I will apply it to hyperv-next later -- I doubt people will object to this change, but just in case.
Thanks,
Wei.
Quoted 50 lines
In addtion, the fuction
hyperv_init()checks ifkmap()fails byBUG_ON(). From the original discussion[2], theBUG_ON()here is just used to explicitly panic NULL pointer. So still keep theBUG_ON()in place to check ifkmap_local_page()fails. Based on this consideration,memcpy_to_page()is not selected here but onlykmap_local_page()is used.Therefore, replace
kmap()withkmap_local_page()inhyperv/hv_init.c.[1]: https://lore.kernel.org/all/[email protected]
[2]: https://lore.kernel.org/lkml/20200915103710.cqmdvzh5lys4wsqo@liuwe-devbox-debian-v2/Suggested-by Dave Hansen <[email protected]>Suggested-by Ira Weiny <[email protected]>Suggested-by Fabio M. De Francesco <[email protected]>Signed-off-by Zhao Liu <[email protected]>---
Suggested by credits.
Dave: Referred to his comments about whetherkmap()can fail and the
suggestion to keepBUG_ON()in place.
Ira: Referred to his task documentation and review comments about
keepingBUG_ON()forkmap_local_page().
Fabio: Stole some of his boiler plate commit message.--- arch/x86/hyperv/hv_init.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)arch/x86/hyperv/hv_init.c +2
-2@@ -459,13 +459,13 @@ void __init hyperv_init(void)wrmsrl (HV_X64_MSR_HYPERCALL , hypercall_msr.as_uint64); pg =vmalloc_to_page (hv_hypercall_pg); - dst =kmap (pg); + dst =kmap_local_page (pg); src =memremap (hypercall_msr.guest_physical_address <<PAGE_SHIFT ,PAGE_SIZE ,MEMREMAP_WB );BUG_ON (!(src && dst));memcpy (dst, src,HV_HYP_PAGE_SIZE );memunmap (src); -kunmap (pg); +kunmap_local (dst); }else { hypercall_msr.guest_physical_address =vmalloc_to_pfn (hv_hypercall_pg);wrmsrl (HV_X64_MSR_HYPERCALL , hypercall_msr.as_uint64);Signature · 1 line
2.34.1
Quoted 27 lines · Wei Liu
Quoted 16 lines · Zhao Liu
From: Zhao Liu <[email protected]>
kmap()is being deprecated in favor of kmap_local_page()[1].There are two main problems with
kmap(): (1) It comes with an overhead as mapping space is restricted and protected by a global lock for synchronization and (2) it also requires global TLB invalidation when the kmap's pool wraps and it might block when the mapping space is fully utilized until a slot becomes available.With
kmap_local_page()the mappings are per thread, CPU local, can take page faults, and can be called from any context (including interrupts). It is faster thankmap()in kernels with HIGHMEM enabled. Furthermore, the tasks can be preempted and, when they are scheduled to run again, the kernel virtual addresses are restored and are still valid.Quoted 5 lines
In the fuction
hyperv_init()ofhyperv/hv_init.c, the mapping is used in a single thread and is short live. So, in this case, it's safe to simply usekmap_local_page()to create mapping, and this avoids the wasted cost ofkmap()for global synchronization.The kmap call in that function is not performance critical in any way, and at this point in the initialization process I don't expect there to be any contention, so the downside of kmap is not really a concern here.
Quoted 6 lines
That being said, kmap getting deprecated is a good enough reason to switch to
kmap_local_page. And I appreciate this well-written, well-reasoned commit message.I will apply it to hyperv-next later -- I doubt people will object to this change, but just in case.
Applied to hyperv-next. Thanks.
Quoted 32 lines · Wei Liu
Date: Mon, 3 Oct 2022 08:50:16 +0000
From: Wei Liu <[email protected]>
Subject: Re: [PATCH] x86/hyperv: Replacekmap()withkmap_local_page()Quoted 27 lines · Wei Liu
Quoted 16 lines · Zhao Liu
From: Zhao Liu <[email protected]>
kmap()is being deprecated in favor of kmap_local_page()[1].There are two main problems with
kmap(): (1) It comes with an overhead as mapping space is restricted and protected by a global lock for synchronization and (2) it also requires global TLB invalidation when the kmap's pool wraps and it might block when the mapping space is fully utilized until a slot becomes available.With
kmap_local_page()the mappings are per thread, CPU local, can take page faults, and can be called from any context (including interrupts). It is faster thankmap()in kernels with HIGHMEM enabled. Furthermore, the tasks can be preempted and, when they are scheduled to run again, the kernel virtual addresses are restored and are still valid.Quoted 5 lines
In the fuction
hyperv_init()ofhyperv/hv_init.c, the mapping is used in a single thread and is short live. So, in this case, it's safe to simply usekmap_local_page()to create mapping, and this avoids the wasted cost ofkmap()for global synchronization.The kmap call in that function is not performance critical in any way, and at this point in the initialization process I don't expect there to be any contention, so the downside of kmap is not really a concern here.
Quoted 8 lines
Quoted 6 lines
That being said, kmap getting deprecated is a good enough reason to switch to
kmap_local_page. And I appreciate this well-written, well-reasoned commit message.I will apply it to hyperv-next later -- I doubt people will object to this change, but just in case.
Applied to hyperv-next. Thanks.
Sorry Wei, based on Ira and Fabio's comments, do you agree me to send a follow on patch to remove that BUG_ON()? Or send the v2 patch?
Thanks,
Zhao
Wei Liu
maintainerQuoted 37 lines · Zhao Liu
Quoted 36 lines · Wei Liu
Date: Mon, 3 Oct 2022 08:50:16 +0000
From: Wei Liu <[email protected]>
Subject: Re: [PATCH] x86/hyperv: Replacekmap()withkmap_local_page()Quoted 31 lines · Wei Liu
Quoted 16 lines · Zhao Liu
From: Zhao Liu <[email protected]>
kmap()is being deprecated in favor of kmap_local_page()[1].There are two main problems with
kmap(): (1) It comes with an overhead as mapping space is restricted and protected by a global lock for synchronization and (2) it also requires global TLB invalidation when the kmap's pool wraps and it might block when the mapping space is fully utilized until a slot becomes available.With
kmap_local_page()the mappings are per thread, CPU local, can take page faults, and can be called from any context (including interrupts). It is faster thankmap()in kernels with HIGHMEM enabled. Furthermore, the tasks can be preempted and, when they are scheduled to run again, the kernel virtual addresses are restored and are still valid.Quoted 5 lines
In the fuction
hyperv_init()ofhyperv/hv_init.c, the mapping is used in a single thread and is short live. So, in this case, it's safe to simply usekmap_local_page()to create mapping, and this avoids the wasted cost ofkmap()for global synchronization.The kmap call in that function is not performance critical in any way, and at this point in the initialization process I don't expect there to be any contention, so the downside of kmap is not really a concern here.
That being said, kmap getting deprecated is a good enough reason to switch to
kmap_local_page. And I appreciate this well-written, well-reasoned commit message.
Quoted 8 lines
Quoted 4 lines
Quoted 2 lines
I will apply it to hyperv-next later -- I doubt people will object to this change, but just in case.
Applied to hyperv-next. Thanks.
Sorry Wei, based on Ira and Fabio's comments, do you agree me to send a follow on patch to remove that
BUG_ON()? Or send the v2 patch?
I just sent a PR to Linus. That PR includes your v1 patch.
While the code can be improved, BUG_ON that does nothing is harmless. You can send a follow-up patch to remove the BUG_ON.
Thanks,
Wei.
Quoted 2 lines
Thanks,
Zhao
Quoted 47 lines · Wei Liu
Date: Sun, 9 Oct 2022 20:46:33 +0000
From: Wei Liu <[email protected]>
Subject: Re: [PATCH] x86/hyperv: Replacekmap()withkmap_local_page()Quoted 42 lines · Zhao Liu
Quoted 40 lines · Wei Liu
Date: Mon, 3 Oct 2022 08:50:16 +0000
From: Wei Liu <[email protected]>
Subject: Re: [PATCH] x86/hyperv: Replacekmap()withkmap_local_page()Quoted 27 lines · Wei Liu
Quoted 16 lines · Zhao Liu
From: Zhao Liu <[email protected]>
kmap()is being deprecated in favor of kmap_local_page()[1].There are two main problems with
kmap(): (1) It comes with an overhead as mapping space is restricted and protected by a global lock for synchronization and (2) it also requires global TLB invalidation when the kmap's pool wraps and it might block when the mapping space is fully utilized until a slot becomes available.With
kmap_local_page()the mappings are per thread, CPU local, can take page faults, and can be called from any context (including interrupts). It is faster thankmap()in kernels with HIGHMEM enabled. Furthermore, the tasks can be preempted and, when they are scheduled to run again, the kernel virtual addresses are restored and are still valid.Quoted 5 lines
In the fuction
hyperv_init()ofhyperv/hv_init.c, the mapping is used in a single thread and is short live. So, in this case, it's safe to simply usekmap_local_page()to create mapping, and this avoids the wasted cost ofkmap()for global synchronization.The kmap call in that function is not performance critical in any way, and at this point in the initialization process I don't expect there to be any contention, so the downside of kmap is not really a concern here.
Quoted 6 lines
That being said, kmap getting deprecated is a good enough reason to switch to
kmap_local_page. And I appreciate this well-written, well-reasoned commit message.I will apply it to hyperv-next later -- I doubt people will object to this change, but just in case.
Applied to hyperv-next. Thanks.
Quoted 8 lines
Quoted 3 lines
Sorry Wei, based on Ira and Fabio's comments, do you agree me to send a follow on patch to remove that
BUG_ON()? Or send the v2 patch?I just sent a PR to Linus. That PR includes your v1 patch.
While the code can be improved,
BUG_ONthat does nothing is harmless. You can send a follow-up patch to remove theBUG_ON.
Thanks Wei! I'll do that.
Regards,
Zhao
Quoted 6 lines
Thanks,
Wei.Quoted 2 lines
Thanks,
Zhao
Quoted 21 lines · CEST Zhao Liu
From: Zhao Liu <[email protected]>
kmap()is being deprecated in favor of kmap_local_page()[1].There are two main problems with
kmap(): (1) It comes with an overhead as mapping space is restricted and protected by a global lock for synchronization and (2) it also requires global TLB invalidation when the kmap's pool wraps and it might block when the mapping space is fully utilized until a slot becomes available.With
kmap_local_page()the mappings are per thread, CPU local, can take page faults, and can be called from any context (including interrupts). It is faster thankmap()in kernels with HIGHMEM enabled. Furthermore, the tasks can be preempted and, when they are scheduled to run again, the kernel virtual addresses are restored and are still valid.In the fuction
hyperv_init()ofhyperv/hv_init.c, the mapping is used in a single thread and is short live. So, in this case, it's safe to simply usekmap_local_page()to create mapping, and this avoids the wasted cost ofkmap()for global synchronization.
Quoted 4 lines
In addtion, the fuction
hyperv_init()checks ifkmap()fails byBUG_ON(). From the original discussion[2], theBUG_ON()here is just used to explicitly panic NULL pointer. So still keep theBUG_ON()in place to check ifkmap_local_page()fails.
How might kmap_local_page() return invalid kernel addresses?
I think that, if this function returns, the pointer is always a valid kernel address. Am I missing something?
Quoted 2 lines
Based on this consideration,
memcpy_to_page()is not selected here but onlykmap_local_page()is used.
I can't agree with you, if the premises are that kmap_local_page() might provide invalid addresses.
Thanks,
Fabio
Quoted 5 lines
Therefore, replace
kmap()withkmap_local_page()inhyperv/hv_init.c.[1]: https://lore.kernel.org/all/[email protected]
[2]: https://lore.kernel.org/lkml/20200915103710.cqmdvzh5lys4wsqo@liuwe-devbox-debian-v2/
Quoted 8 lines
Suggested-by Dave Hansen <[email protected]>Suggested-by Ira Weiny <[email protected]>Suggested-by Fabio M. De Francesco <[email protected]>Signed-off-by Zhao Liu <[email protected]>---
Suggested by credits.
Dave: Referred to his comments about whetherkmap()can fail and
the
Quoted 8 lines
suggestion to keep
BUG_ON()in place.
Ira: Referred to his task documentation and review comments about
keepingBUG_ON()forkmap_local_page().
Fabio: Stole some of his boiler plate commit message.--- arch/x86/hyperv/hv_init.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
Quoted 27 lines · Fabio M. De Francesco
Quoted 21 lines · CEST Zhao Liu
From: Zhao Liu <[email protected]>
kmap()is being deprecated in favor of kmap_local_page()[1].There are two main problems with
kmap(): (1) It comes with an overhead as mapping space is restricted and protected by a global lock for synchronization and (2) it also requires global TLB invalidation when the kmap's pool wraps and it might block when the mapping space is fully utilized until a slot becomes available.With
kmap_local_page()the mappings are per thread, CPU local, can take page faults, and can be called from any context (including interrupts). It is faster thankmap()in kernels with HIGHMEM enabled. Furthermore, the tasks can be preempted and, when they are scheduled to run again, the kernel virtual addresses are restored and are still valid.In the fuction
hyperv_init()ofhyperv/hv_init.c, the mapping is used in a single thread and is short live. So, in this case, it's safe to simply usekmap_local_page()to create mapping, and this avoids the wasted cost ofkmap()for global synchronization.Quoted 4 lines
In addtion, the fuction
hyperv_init()checks ifkmap()fails byBUG_ON(). From the original discussion[2], theBUG_ON()here is just used to explicitly panic NULL pointer. So still keep theBUG_ON()in place to check ifkmap_local_page()fails.
Quoted 4 lines
How might
kmap_local_page()return invalid kernel addresses?I think that, if this function returns, the pointer is always a valid kernel address. Am I missing something?
I think this is my mistake. I did not realize what Zhao was asking me the other day and generally said to leave BUG_ON's alone and not change things.
In this case kmap_local_page() will not return NULL. It will BUG on its own if it fails.
Quoted 6 lines
Quoted 3 lines
Based on this consideration,
memcpy_to_page()is not selected here but onlykmap_local_page()is used.I can't agree with you, if the premises are that
kmap_local_page()might provide invalid addresses.
I'm ok with the patch as is. But Fabio is correct and it may be worth removing the check in a follow on patch.
Ira
Quoted 29 lines
Thanks,
Fabio
Quoted 5 lines
Therefore, replace
kmap()withkmap_local_page()inhyperv/hv_init.c.[1]: https://lore.kernel.org/all/[email protected]
[2]: https://lore.kernel.org/lkml/20200915103710.cqmdvzh5lys4wsqo@liuwe-devbox-debian-v2/Quoted 8 lines
Suggested-by Dave Hansen <[email protected]>Suggested-by Ira Weiny <[email protected]>Suggested-by Fabio M. De Francesco <[email protected]>Signed-off-by Zhao Liu <[email protected]>---
Suggested by credits.
Dave: Referred to his comments about whetherkmap()can fail andthe
Quoted 8 lines
suggestion to keep
BUG_ON()in place.
Ira: Referred to his task documentation and review comments about
keepingBUG_ON()forkmap_local_page().
Fabio: Stole some of his boiler plate commit message.--- arch/x86/hyperv/hv_init.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
Quoted 32 lines · Ira Weiny
Date: Mon, 3 Oct 2022 13:24:51 -0700
From: Ira Weiny <[email protected]>
Subject: Re: [PATCH] x86/hyperv: Replacekmap()withkmap_local_page()Quoted 27 lines · Fabio M. De Francesco
Quoted 25 lines · CEST Zhao Liu
From: Zhao Liu <[email protected]>
kmap()is being deprecated in favor of kmap_local_page()[1].There are two main problems with
kmap(): (1) It comes with an overhead as mapping space is restricted and protected by a global lock for synchronization and (2) it also requires global TLB invalidation when the kmap's pool wraps and it might block when the mapping space is fully utilized until a slot becomes available.With
kmap_local_page()the mappings are per thread, CPU local, can take page faults, and can be called from any context (including interrupts). It is faster thankmap()in kernels with HIGHMEM enabled. Furthermore, the tasks can be preempted and, when they are scheduled to run again, the kernel virtual addresses are restored and are still valid.In the fuction
hyperv_init()ofhyperv/hv_init.c, the mapping is used in a single thread and is short live. So, in this case, it's safe to simply usekmap_local_page()to create mapping, and this avoids the wasted cost ofkmap()for global synchronization.In addtion, the fuction
hyperv_init()checks ifkmap()fails byBUG_ON(). From the original discussion[2], theBUG_ON()here is just used to explicitly panic NULL pointer. So still keep theBUG_ON()in place to check ifkmap_local_page()fails.
Quoted 7 lines
Quoted 4 lines
How might
kmap_local_page()return invalid kernel addresses?I think that, if this function returns, the pointer is always a valid kernel address. Am I missing something?
I think this is my mistake. I did not realize what Zhao was asking me the other day and generally said to leave BUG_ON's alone and not change things.
Thanks Fabio and Ira! This is my negligence, my consideration is still not thorough enough. I'll remove this BUG_ON().
Quoted 44 lines
In this case
kmap_local_page()will not return NULL. It will BUG on its own if it fails.Quoted 6 lines
Quoted 3 lines
Based on this consideration,
memcpy_to_page()is not selected here but onlykmap_local_page()is used.I can't agree with you, if the premises are that
kmap_local_page()might provide invalid addresses.I'm ok with the patch as is. But Fabio is correct and it may be worth removing the check in a follow on patch.
IraQuoted 29 lines
Thanks,
Fabio
Quoted 5 lines
Therefore, replace
kmap()withkmap_local_page()inhyperv/hv_init.c.[1]: https://lore.kernel.org/all/[email protected]
[2]: https://lore.kernel.org/lkml/20200915103710.cqmdvzh5lys4wsqo@liuwe-devbox-debian-v2/Quoted 8 lines
Suggested-by Dave Hansen <[email protected]>Suggested-by Ira Weiny <[email protected]>Suggested-by Fabio M. De Francesco <[email protected]>Signed-off-by Zhao Liu <[email protected]>---
Suggested by credits.
Dave: Referred to his comments about whetherkmap()can fail andthe
Quoted 8 lines
suggestion to keep
BUG_ON()in place.
Ira: Referred to his task documentation and review comments about
keepingBUG_ON()forkmap_local_page().
Fabio: Stole some of his boiler plate commit message.--- arch/x86/hyperv/hv_init.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)