linux-kernel - [PATCH] io-mapping: Don't disable preempt on RT in io_mapping_map_atomic

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>] [day] [month] [year] [list]

Message-ID: <20230310162905.O57Pj7hh@linutronix.de>
Date:   Fri, 10 Mar 2023 17:29:05 +0100
From:   Sebastian Andrzej Siewior <bigeasy@...utronix.de>
To:     linux-kernel@...r.kernel.org, linux-mm@...ck.org
Cc:     Andrew Morton <akpm@...ux-foundation.org>,
        Thomas Gleixner <tglx@...utronix.de>,
        Richard Weinberger <richard.weinberger@...il.com>
Subject: [PATCH] io-mapping: Don't disable preempt on RT in
 io_mapping_map_atomic_wc().

io_mapping_map_atomic_wc() disables preemption and pagefaults for historical
reasons. The conversion to io_mapping_map_local_wc(), which only disables
migration, cannot be done wholesale because quite some call sites need to be
updated to accommodate with the changed semantics.

On PREEMPT_RT enabled kernels the io_mapping_map_atomic_wc() semantics are
problematic due to the implicit disabling of preemption which makes it
impossible to acquire 'sleeping' spinlocks within the mapped atomic sections.

PREEMPT_RT replaces the preempt_disable() with a migrate_disable() for
more than a decade. It could be argued that this is a justification to do
this unconditionally, but PREEMPT_RT covers only a limited number of
architectures and it disables some functionality which limits the coverage
further.

Limit the replacement to PREEMPT_RT for now. This is also done
kmap_atomic().

Reported-by: Richard Weinberger <richard.weinberger@...il.com>
Link: https://lore.kernel.org/CAFLxGvw0WMxaMqYqJ5WgvVSbKHq2D2xcXTOgMCpgq9nDC-MWTQ@mail.gmail.com
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@...utronix.de>
---

Richard reported it against the i915 driver. I got offlist another
report and this reporter confirmed that it fixes the issue while no
glitches or other problems were observed (in the kernel log and on
screen).

 include/linux/io-mapping.h | 20 ++++++++++++++++----
 1 file changed, 16 insertions(+), 4 deletions(-)

diff --git a/include/linux/io-mapping.h b/include/linux/io-mapping.h
index 09d4f17c8d3b6..7376c1df9c901 100644
--- a/include/linux/io-mapping.h
+++ b/include/linux/io-mapping.h
@@ -69,7 +69,10 @@ io_mapping_map_atomic_wc(struct io_mapping *mapping,

 	BUG_ON(offset >= mapping->size);
 	phys_addr = mapping->base + offset;
-	preempt_disable();
+	if (!IS_ENABLED(CONFIG_PREEMPT_RT))
+		preempt_disable();
+	else
+		migrate_disable();
 	pagefault_disable();
 	return __iomap_local_pfn_prot(PHYS_PFN(phys_addr), mapping->prot);
 }
@@ -79,7 +82,10 @@ io_mapping_unmap_atomic(void __iomem *vaddr)
 {
 	kunmap_local_indexed((void __force *)vaddr);
 	pagefault_enable();
-	preempt_enable();
+	if (!IS_ENABLED(CONFIG_PREEMPT_RT))
+		preempt_enable();
+	else
+		migrate_enable();
 }

 static inline void __iomem *
@@ -162,7 +168,10 @@ static inline void __iomem *
 io_mapping_map_atomic_wc(struct io_mapping *mapping,
 			 unsigned long offset)
 {
-	preempt_disable();
+	if (!IS_ENABLED(CONFIG_PREEMPT_RT))
+		preempt_disable();
+	else
+		migrate_disable();
 	pagefault_disable();
 	return io_mapping_map_wc(mapping, offset, PAGE_SIZE);
 }
@@ -172,7 +181,10 @@ io_mapping_unmap_atomic(void __iomem *vaddr)
 {
 	io_mapping_unmap(vaddr);
 	pagefault_enable();
-	preempt_enable();
+	if (!IS_ENABLED(CONFIG_PREEMPT_RT))
+		preempt_enable();
+	else
+		migrate_enable();
 }

 static inline void __iomem *
-- 
2.39.2