Skip to contentSkip to conversation

powerpc/svm: Don't issue ultracalls if !mem_encrypt_active()

PATCH In review Will Deacon 4 participants 4 messages Jul 30, 2021 – Aug 13, 2021 PowerPC

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 powerpc/svm: Don't issue ultracalls if !mem_encrypt_active()

Commit ad6c00283163 ("swiotlb: Free tbl memory in swiotlb_exit()") introduced a set_memory_encrypted() call to swiotlb_exit() so that the buffer pages are returned to an encrypted state prior to being freed.

Sachin reports that this leads to the following crash on a Power server:

[    0.010799] software IO TLB: tearing down default memory pool
[    0.010805] ------------[ cut here ]------------
[    0.010808] kernel BUG at arch/powerpc/kernel/interrupt.c:98!

Nick spotted that this is because set_memory_encrypted() is issuing an ultracall which doesn't exist for the processor, and should therefore be gated by mem_encrypt_active() to mirror the x86 implementation.

Cc Konrad Rzeszutek Wilk <[email protected]>
Cc Claire Chang <[email protected]>
Cc Christoph Hellwig <[email protected]>
Cc Robin Murphy <[email protected]>
Fixes ad6c00283163 ("swiotlb: Free tbl memory in swiotlb_exit()")
---
 arch/powerpc/platforms/pseries/svm.c | 6 ++++++
 1 file changed, 6 insertions(+)
arch/powerpc/platforms/pseries/svm.c +6 -0
index 1d829e257996..87f001b4c4e4 100644
@@ -63,6 +63,9 @@ void __init svm_swiotlb_init(void)
 
 int set_memory_encrypted(unsigned long addr, int numpages)
 {
+	if (!mem_encrypt_active())
+		return 0;
+
 	if (!PAGE_ALIGNED(addr))
 		return -EINVAL;
 
@@ -73,6 +76,9 @@ int set_memory_encrypted(unsigned long addr, int numpages)
 
 int set_memory_decrypted(unsigned long addr, int numpages)
 {
+	if (!mem_encrypt_active())
+		return 0;
+
 	if (!PAGE_ALIGNED(addr))
 		return -EINVAL;
 
Signature · 1 line
2.32.0.554.ge1b32706d8-goog
reviewer Acked-by on v1 1/1
#1 lore raw

Excerpts from Will Deacon's message of July 30, 2021 9:42 pm:

Quoted 25 lines

Commit ad6c00283163 ("swiotlb: Free tbl memory in swiotlb_exit()") introduced a set_memory_encrypted() call to swiotlb_exit() so that the buffer pages are returned to an encrypted state prior to being freed.

Sachin reports that this leads to the following crash on a Power server:

[    0.010799] software IO TLB: tearing down default memory pool
[    0.010805] ------------[ cut here ]------------
[    0.010808] kernel BUG at arch/powerpc/kernel/interrupt.c:98!

Nick spotted that this is because set_memory_encrypted() is issuing an ultracall which doesn't exist for the processor, and should therefore be gated by mem_encrypt_active() to mirror the x86 implementation.

Cc Konrad Rzeszutek Wilk <[email protected]>
Cc Claire Chang <[email protected]>
Cc Christoph Hellwig <[email protected]>
Cc Robin Murphy <[email protected]>
Fixes ad6c00283163 ("swiotlb: Free tbl memory in swiotlb_exit()")
Quoted 1 line

---

Thanks for writing it.

Quoted 31 lines
 arch/powerpc/platforms/pseries/svm.c | 6 ++++++
 1 file changed, 6 insertions(+)
arch/powerpc/platforms/pseries/svm.c +6 -0
index 1d829e257996..87f001b4c4e4 100644
@@ -63,6 +63,9 @@ void __init svm_swiotlb_init(void)
 
 int set_memory_encrypted(unsigned long addr, int numpages)
 {
+	if (!mem_encrypt_active())
+		return 0;
+
 	if (!PAGE_ALIGNED(addr))
 		return -EINVAL;
 
@@ -73,6 +76,9 @@ int set_memory_encrypted(unsigned long addr, int numpages)
 
 int set_memory_decrypted(unsigned long addr, int numpages)
 {
+	if (!mem_encrypt_active())
+		return 0;
+
 	if (!PAGE_ALIGNED(addr))
 		return -EINVAL;
 
Signature · 1 line
2.32.0.554.ge1b32706d8-goog

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu

reviewer Acked-by on v1 1/1
#2 lore raw
In reply to Will Deacon
Quoted 25 lines · Will Deacon

Commit ad6c00283163 ("swiotlb: Free tbl memory in swiotlb_exit()") introduced a set_memory_encrypted() call to swiotlb_exit() so that the buffer pages are returned to an encrypted state prior to being freed.

Sachin reports that this leads to the following crash on a Power server:

[    0.010799] software IO TLB: tearing down default memory pool
[    0.010805] ------------[ cut here ]------------
[    0.010808] kernel BUG at arch/powerpc/kernel/interrupt.c:98!

Nick spotted that this is because set_memory_encrypted() is issuing an ultracall which doesn't exist for the processor, and should therefore be gated by mem_encrypt_active() to mirror the x86 implementation.

Cc Konrad Rzeszutek Wilk <[email protected]>
Cc Claire Chang <[email protected]>
Cc Christoph Hellwig <[email protected]>
Cc Robin Murphy <[email protected]>
Fixes ad6c00283163 ("swiotlb: Free tbl memory in swiotlb_exit()")
Quoted 3 lines
---
 arch/powerpc/platforms/pseries/svm.c | 6 ++++++
 1 file changed, 6 insertions(+)

Thanks.

I assume Konrad will take this via the swiotlb tree?

cheers

#3 lore raw
Quoted 30 lines · Michael Ellerman
Quoted 25 lines · Will Deacon

Commit ad6c00283163 ("swiotlb: Free tbl memory in swiotlb_exit()") introduced a set_memory_encrypted() call to swiotlb_exit() so that the buffer pages are returned to an encrypted state prior to being freed.

Sachin reports that this leads to the following crash on a Power server:

[    0.010799] software IO TLB: tearing down default memory pool
[    0.010805] ------------[ cut here ]------------
[    0.010808] kernel BUG at arch/powerpc/kernel/interrupt.c:98!

Nick spotted that this is because set_memory_encrypted() is issuing an ultracall which doesn't exist for the processor, and should therefore be gated by mem_encrypt_active() to mirror the x86 implementation.

Cc Konrad Rzeszutek Wilk <[email protected]>
Cc Claire Chang <[email protected]>
Cc Christoph Hellwig <[email protected]>
Cc Robin Murphy <[email protected]>
Fixes ad6c00283163 ("swiotlb: Free tbl memory in swiotlb_exit()")
Quoted 3 lines
---
 arch/powerpc/platforms/pseries/svm.c | 6 ++++++
 1 file changed, 6 insertions(+)
Quoted 6 lines

Thanks.

I assume Konrad will take this via the swiotlb tree?

Btw, we're currently reworking that whole "am I running as a confidential guest" querying, see:

https://lkml.kernel.org/r/029791b24c6412f9427cfe6ec598156c64395964.1627424774.git.thomas.lendacky@amd.com

for example.

I see Konrad has queued this for 5.15 in his devel/for-linus-5.15 branch so if he sends it to Linus in the upcoming merge window (right Konrad?) then I can base the rework ontop, once 5.15-rc1 releases, so that there are no build breakages...

Thx.

Signature · 4 lines
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette