mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Uros Bizjak <ubizjak@gmail.com>
To: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>,
	linux-kernel@vger.kernel.org, iommu@lists.linux.dev
Cc: joro@8bytes.org, robin.murphy@arm.com, vasant.hegde@amd.com,
	jgg@nvidia.com, kevin.tian@intel.com, jon.grimm@amd.com,
	santosh.shukla@amd.com, pandoh@google.com, kumaranand@google.com
Subject: Re: [PATCH v5 2/6] iommu/amd: Introduce helper function to update 256-bit DTE
Date: Mon, 7 Oct 2024 16:42:30 +0200	[thread overview]
Message-ID: <cc44a5ca-ddf9-ca7a-93f8-38bf26ac1f1f@gmail.com> (raw)
In-Reply-To: <20241007041353.4756-3-suravee.suthikulpanit@amd.com>

[-- Attachment #1: Type: text/plain, Size: 2012 bytes --]



On 7. 10. 24 06:13, Suravee Suthikulpanit wrote:

> +
>   /****************************************************************************
>    *
>    * Helper functions
>    *
>    ****************************************************************************/
>   
> +static void write_dte_upper128(struct dev_table_entry *ptr, struct dev_table_entry *new)
> +{
> +	struct dev_table_entry old = {};
> +
> +	do {
> +		old.data128[1] = ptr->data128[1];
> +		new->data[2] &= ~DTE_DATA2_INTR_MASK;
> +		new->data[2] |= old.data[2] & (DTE_DATA2_INTR_MASK | DTE_DATA2_RESV_MASK);
> +	} while (!try_cmpxchg128(&ptr->data128[1], &old.data128[1], new->data128[1]));

Please note that try_cmpxchg inherently updates &old.data128[1] above on 
failure. There is no need to update value again in the loop.

Please also note that the value from ptr->data128[1] should be read 
using READ_ONCE() to prevent compiler from merging, refetching or 
reordering the read. Currently, there is no READ_ONCE() implemented for 
__int128, so something like the attached patch should be used.

Based on the above, the loop should be rewritten as:

	old.data128[1] = READ_ONCE(ptr->data128[1]);
	do {
		new->data[2] &= ~DTE_DATA2_INTR_MASK;
		new->data[2] |= old.data[2] & (DTE_DATA2_INTR_MASK | DTE_DATA2_RESV_MASK);
	} while (!try_cmpxchg128(&ptr->data128[1], &old.data128[1], 
new->data128[1]));

> +}
> +
> +static void write_dte_lower128(struct dev_table_entry *ptr, struct dev_table_entry *new)
> +{
> +	struct dev_table_entry old = {};
> +
> +	/*
> +	 * Need to preserve DTE[96:106], which can be set by information in IVRS table.
> +	 * See set_dev_entry_from_acpi().
> +	 */
> +	new->data[1] |= ptr->data[1] & DTE_FLAG_MASK;
> +
> +	do {
> +		old.data128[0] = ptr->data128[0];
> +	} while (!try_cmpxchg128(&ptr->data128[0], &old.data128[0], new->data128[0]));

And this one as:

	old.data128[0] = READ_ONCE(ptr->data128[0]);
	do {
	} while (!try_cmpxchg128(&ptr->data128[0], &old.data128[0], 
new->data128[0]));

Best regards,
Uros.

[-- Attachment #2: p.diff.txt --]
[-- Type: text/plain, Size: 1411 bytes --]

diff --git a/include/asm-generic/rwonce.h b/include/asm-generic/rwonce.h
index 8d0a6280e982..8bf942ad5ef3 100644
--- a/include/asm-generic/rwonce.h
+++ b/include/asm-generic/rwonce.h
@@ -33,7 +33,7 @@
  * (e.g. a virtual address) and a strong prevailing wind.
  */
 #define compiletime_assert_rwonce_type(t)					\
-	compiletime_assert(__native_word(t) || sizeof(t) == sizeof(long long),	\
+	compiletime_assert(__native_word(t) || sizeof(t) == sizeof(__dword_type), \
 		"Unsupported access size for {READ,WRITE}_ONCE().")
 
 /*
diff --git a/include/linux/compiler_types.h b/include/linux/compiler_types.h
index 94b8fedfb077..8615e91f48fd 100644
--- a/include/linux/compiler_types.h
+++ b/include/linux/compiler_types.h
@@ -469,6 +469,12 @@ struct ftrace_likely_data {
 		unsigned type:	(unsigned type)0,			\
 		signed type:	(signed type)0
 
+#ifdef __SIZEOF_INT128__
+#define __dword_type __int128
+#else
+#define __dword_type long long
+#endif
+
 #define __unqual_scalar_typeof(x) typeof(				\
 		_Generic((x),						\
 			 char:	(char)0,				\
@@ -476,7 +482,7 @@ struct ftrace_likely_data {
 			 __scalar_type_to_expr_cases(short),		\
 			 __scalar_type_to_expr_cases(int),		\
 			 __scalar_type_to_expr_cases(long),		\
-			 __scalar_type_to_expr_cases(long long),	\
+			 __scalar_type_to_expr_cases(__dword_type),	\
 			 default: (x)))
 
 /* Is this type a native word size -- useful for atomic operations */

  parent reply	other threads:[~2024-10-07 14:42 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-10-07  4:13 [PATCH v5 0/6] iommu/amd: Use 128-bit cmpxchg operation to update DTE Suravee Suthikulpanit
2024-10-07  4:13 ` [PATCH v5 1/6] iommu/amd: Disable AMD IOMMU if CMPXCHG16B feature is not supported Suravee Suthikulpanit
2024-10-07  4:13 ` [PATCH v5 2/6] iommu/amd: Introduce helper function to update 256-bit DTE Suravee Suthikulpanit
2024-10-07 14:06   ` Jason Gunthorpe
2024-10-11  5:25     ` Suthikulpanit, Suravee
2024-10-07 14:42   ` Uros Bizjak [this message]
2024-10-11 10:22     ` Suthikulpanit, Suravee
2024-10-11 11:05       ` Uros Bizjak
2024-10-07  4:13 ` [PATCH v5 3/6] iommu/amd: Modify set_dte_entry() to use 256-bit DTE helpers Suravee Suthikulpanit
2024-10-07 14:05   ` Jason Gunthorpe
2024-10-16  5:12     ` Suthikulpanit, Suravee
2024-10-07 14:17   ` Jason Gunthorpe
2024-10-16  5:15     ` Suthikulpanit, Suravee
2024-10-07  4:13 ` [PATCH v5 4/6] iommu/amd: Introduce helper function get_dte256() Suravee Suthikulpanit
2024-10-07  4:13 ` [PATCH v5 5/6] iommu/amd: Modify clear_dte_entry() to avoid in-place update Suravee Suthikulpanit
2024-10-07 14:10   ` Jason Gunthorpe
2024-10-07  4:13 ` [PATCH v5 6/6] iommu/amd: Lock DTE before updating the entry with WRITE_ONCE() Suravee Suthikulpanit

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=cc44a5ca-ddf9-ca7a-93f8-38bf26ac1f1f@gmail.com \
    --to=ubizjak@gmail.com \
    --cc=iommu@lists.linux.dev \
    --cc=jgg@nvidia.com \
    --cc=jon.grimm@amd.com \
    --cc=joro@8bytes.org \
    --cc=kevin.tian@intel.com \
    --cc=kumaranand@google.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=pandoh@google.com \
    --cc=robin.murphy@arm.com \
    --cc=santosh.shukla@amd.com \
    --cc=suravee.suthikulpanit@amd.com \
    --cc=vasant.hegde@amd.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

Powered by JetHome