LKML Archive on
help / color / mirror / Atom feed
From: Logan Gunthorpe <>
	Andrew Morton <>
Cc: "Arnd Bergmann" <>,
	"Greg Kroah-Hartman" <>,
	"Andy Shevchenko" <>,
	"Horia Geantă" <>,
	"Logan Gunthorpe" <>,
	"Thomas Gleixner" <>,
	"Kate Stewart" <>,
	"Philippe Ombredanne" <>
Subject: [PATCH v17 1/7] iomap: Use non-raw io functions for io{read|write}XXbe
Date: Mon,  7 May 2018 14:35:59 -0600	[thread overview]
Message-ID: <> (raw)
In-Reply-To: <>

Fix an asymmetry in the io{read|write}XXbe functions in that the
big-endian variants make use of the raw io accessors while the
little-endian variants use the regular accessors. Some architectures
implement barriers to order against both spinlocks and DMA accesses
and for these case, the big-endian variant of the API would not be

Thus, change the mmio_XXXXbe macros to use the appropriate swab() function
wrapping the regular accessor. This is similar to what was done for PIO.

When this code was originally written, barriers in the IO accessors were
not common and the accessors simply wrapped the raw functions in a
conversion to CPU endianness. Since then, barriers have been added in
some architectures and are now missing in the big endian variant of the

This also manages to silence a few sparse warnings that check
for using the correct endian types which the original code did
not annotate correctly.

Signed-off-by: Logan Gunthorpe <>
Cc: Thomas Gleixner <>
Cc: Kate Stewart <>
Cc: Philippe Ombredanne <>
Cc: Greg Kroah-Hartman <>
Cc: Arnd Bergmann <>
 lib/iomap.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/lib/iomap.c b/lib/iomap.c
index 541d926da95e..2c293b22569f 100644
--- a/lib/iomap.c
+++ b/lib/iomap.c
@@ -65,8 +65,8 @@ static void bad_io_access(unsigned long port, const char *access)
 #ifndef mmio_read16be
-#define mmio_read16be(addr) be16_to_cpu(__raw_readw(addr))
-#define mmio_read32be(addr) be32_to_cpu(__raw_readl(addr))
+#define mmio_read16be(addr) swab16(readw(addr))
+#define mmio_read32be(addr) swab32(readl(addr))
 unsigned int ioread8(void __iomem *addr)
@@ -106,8 +106,8 @@ EXPORT_SYMBOL(ioread32be);
 #ifndef mmio_write16be
-#define mmio_write16be(val,port) __raw_writew(be16_to_cpu(val),port)
-#define mmio_write32be(val,port) __raw_writel(be32_to_cpu(val),port)
+#define mmio_write16be(val,port) writew(swab16(val),port)
+#define mmio_write32be(val,port) writel(swab32(val),port)
 void iowrite8(u8 val, void __iomem *addr)

  reply	other threads:[~2018-05-07 20:35 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-05-07 20:35 [PATCH v17 0/7] Add io{read|write}64 to io-64-atomic headers Logan Gunthorpe
2018-05-07 20:35 ` Logan Gunthorpe [this message]
2018-05-07 20:36 ` [PATCH v17 2/7] parisc: iomap: introduce io{read|write}64 Logan Gunthorpe
2018-05-07 20:36 ` [PATCH v17 3/7] iomap: introduce io{read|write}64_{lo_hi|hi_lo} Logan Gunthorpe
2018-05-07 20:36 ` [PATCH v17 4/7] io-64-nonatomic: add io{read|write}64[be]{_lo_hi|_hi_lo} macros Logan Gunthorpe
2018-05-07 20:36 ` [PATCH v17 5/7] ntb: ntb_hw_intel: use io-64-nonatomic instead of in-driver hacks Logan Gunthorpe
2018-05-07 20:36 ` [PATCH v17 6/7] crypto: caam: cleanup CONFIG_64BIT ifdefs when using io{read|write}64 Logan Gunthorpe
2018-05-07 20:36 ` [PATCH v17 7/7] ntb: ntb_hw_switchtec: Cleanup 64bit IO defines to use the common header Logan Gunthorpe

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:

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

  git send-email \ \ \ \ \ \ \ \ \ \ \ \ \ \ \
    --subject='Re: [PATCH v17 1/7] iomap: Use non-raw io functions for io{read|write}XXbe' \

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).