lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date:   Fri, 02 Dec 2022 14:04:25 -0000
From:   "tip-bot2 for Kristen Carlson Accardi" <tip-bot2@...utronix.de>
To:     linux-tip-commits@...r.kernel.org
Cc:     Kristen Carlson Accardi <kristen@...ux.intel.com>,
        Thomas Gleixner <tglx@...utronix.de>,
        Jarkko Sakkinen <jarkko@...nel.org>,
        "Fabio M. De Francesco" <fmdefrancesco@...il.com>,
        Ira Weiny <ira.weiny@...el.com>, x86@...nel.org,
        linux-kernel@...r.kernel.org
Subject: [tip: x86/sgx] x86/sgx: Replace kmap/kunmap_atomic() calls

The following commit has been merged into the x86/sgx branch of tip:

Commit-ID:     89e927bbcd45d507e5612ef72fda04182e544a38
Gitweb:        https://git.kernel.org/tip/89e927bbcd45d507e5612ef72fda04182e544a38
Author:        Kristen Carlson Accardi <kristen@...ux.intel.com>
AuthorDate:    Tue, 15 Nov 2022 08:16:26 -08:00
Committer:     Thomas Gleixner <tglx@...utronix.de>
CommitterDate: Fri, 02 Dec 2022 14:59:56 +01:00

x86/sgx: Replace kmap/kunmap_atomic() calls

kmap_local_page() is the preferred way to create temporary mappings when it
is feasible, because the mappings are thread-local and CPU-local.

kmap_local_page() uses per-task maps rather than per-CPU maps. This in
effect removes the need to disable preemption on the local CPU while the
mapping is active, and thus vastly reduces overall system latency. It is
also valid to take pagefaults within the mapped region.

The use of kmap_atomic() in the SGX code was not an explicit design choice
to disable page faults or preemption, and there is no compelling design
reason to using kmap_atomic() vs. kmap_local_page().

Signed-off-by: Kristen Carlson Accardi <kristen@...ux.intel.com>
Signed-off-by: Thomas Gleixner <tglx@...utronix.de>
Reviewed-by: Jarkko Sakkinen <jarkko@...nel.org>
Reviewed-by: Fabio M. De Francesco <fmdefrancesco@...il.com>
Reviewed-by: Ira Weiny <ira.weiny@...el.com>
Link: https://lore.kernel.org/linux-sgx/Y0biN3%2FJsZMa0yUr@kernel.org/
Link: https://lore.kernel.org/r/20221115161627.4169428-1-kristen@linux.intel.com
---
 arch/x86/kernel/cpu/sgx/encl.c  | 12 ++++++------
 arch/x86/kernel/cpu/sgx/ioctl.c |  4 ++--
 arch/x86/kernel/cpu/sgx/main.c  |  8 ++++----
 3 files changed, 12 insertions(+), 12 deletions(-)

diff --git a/arch/x86/kernel/cpu/sgx/encl.c b/arch/x86/kernel/cpu/sgx/encl.c
index 2c25825..68f8b18 100644
--- a/arch/x86/kernel/cpu/sgx/encl.c
+++ b/arch/x86/kernel/cpu/sgx/encl.c
@@ -160,8 +160,8 @@ static int __sgx_encl_eldu(struct sgx_encl_page *encl_page,
 		return ret;
 
 	pginfo.addr = encl_page->desc & PAGE_MASK;
-	pginfo.contents = (unsigned long)kmap_atomic(b.contents);
-	pcmd_page = kmap_atomic(b.pcmd);
+	pginfo.contents = (unsigned long)kmap_local_page(b.contents);
+	pcmd_page = kmap_local_page(b.pcmd);
 	pginfo.metadata = (unsigned long)pcmd_page + b.pcmd_offset;
 
 	if (secs_page)
@@ -187,8 +187,8 @@ static int __sgx_encl_eldu(struct sgx_encl_page *encl_page,
 	 */
 	pcmd_page_empty = !memchr_inv(pcmd_page, 0, PAGE_SIZE);
 
-	kunmap_atomic(pcmd_page);
-	kunmap_atomic((void *)(unsigned long)pginfo.contents);
+	kunmap_local(pcmd_page);
+	kunmap_local((void *)(unsigned long)pginfo.contents);
 
 	get_page(b.pcmd);
 	sgx_encl_put_backing(&b);
@@ -197,10 +197,10 @@ static int __sgx_encl_eldu(struct sgx_encl_page *encl_page,
 
 	if (pcmd_page_empty && !reclaimer_writing_to_pcmd(encl, pcmd_first_page)) {
 		sgx_encl_truncate_backing_page(encl, PFN_DOWN(page_pcmd_off));
-		pcmd_page = kmap_atomic(b.pcmd);
+		pcmd_page = kmap_local_page(b.pcmd);
 		if (memchr_inv(pcmd_page, 0, PAGE_SIZE))
 			pr_warn("PCMD page not empty after truncate.\n");
-		kunmap_atomic(pcmd_page);
+		kunmap_local(pcmd_page);
 	}
 
 	put_page(b.pcmd);
diff --git a/arch/x86/kernel/cpu/sgx/ioctl.c b/arch/x86/kernel/cpu/sgx/ioctl.c
index ef87482..03c50f1 100644
--- a/arch/x86/kernel/cpu/sgx/ioctl.c
+++ b/arch/x86/kernel/cpu/sgx/ioctl.c
@@ -221,11 +221,11 @@ static int __sgx_encl_add_page(struct sgx_encl *encl,
 	pginfo.secs = (unsigned long)sgx_get_epc_virt_addr(encl->secs.epc_page);
 	pginfo.addr = encl_page->desc & PAGE_MASK;
 	pginfo.metadata = (unsigned long)secinfo;
-	pginfo.contents = (unsigned long)kmap_atomic(src_page);
+	pginfo.contents = (unsigned long)kmap_local_page(src_page);
 
 	ret = __eadd(&pginfo, sgx_get_epc_virt_addr(epc_page));
 
-	kunmap_atomic((void *)pginfo.contents);
+	kunmap_local((void *)pginfo.contents);
 	put_page(src_page);
 
 	return ret ? -EIO : 0;
diff --git a/arch/x86/kernel/cpu/sgx/main.c b/arch/x86/kernel/cpu/sgx/main.c
index 0aad028..e5a37b6 100644
--- a/arch/x86/kernel/cpu/sgx/main.c
+++ b/arch/x86/kernel/cpu/sgx/main.c
@@ -165,17 +165,17 @@ static int __sgx_encl_ewb(struct sgx_epc_page *epc_page, void *va_slot,
 	pginfo.addr = 0;
 	pginfo.secs = 0;
 
-	pginfo.contents = (unsigned long)kmap_atomic(backing->contents);
-	pginfo.metadata = (unsigned long)kmap_atomic(backing->pcmd) +
+	pginfo.contents = (unsigned long)kmap_local_page(backing->contents);
+	pginfo.metadata = (unsigned long)kmap_local_page(backing->pcmd) +
 			  backing->pcmd_offset;
 
 	ret = __ewb(&pginfo, sgx_get_epc_virt_addr(epc_page), va_slot);
 	set_page_dirty(backing->pcmd);
 	set_page_dirty(backing->contents);
 
-	kunmap_atomic((void *)(unsigned long)(pginfo.metadata -
+	kunmap_local((void *)(unsigned long)(pginfo.metadata -
 					      backing->pcmd_offset));
-	kunmap_atomic((void *)(unsigned long)pginfo.contents);
+	kunmap_local((void *)(unsigned long)pginfo.contents);
 
 	return ret;
 }

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ