Skip to contentSkip to conversation

ata-sff: always map page before data transfer

PATCH Applied Tycho Andersen 3 participants 3 messages May 4, 2017 – May 5, 2017 libata

lore
Share & tools
lore

Tycho Andersen

author
#0 lore raw

PATCH v2 1/1 ata-sff: always map page before data transfer

The XPFO [1] patchset may unmap pages from physmap if they happened to be destined for userspace. If such a page is unmapped, it needs to be remapped. Rather than test if a page is in the highmem/xpfo unmapped state, Christoph suggested [2] that we simply always map the page.

v2: * drop comment about bounce buffer
* don't save IRQs before kmap/unmap
* formatting

Suggested-by Christoph Hellwig <[email protected]>
Signed-off-by Tycho Andersen <[email protected]>
CC Juerg Haefliger <[email protected]>
CC Tejun Heo <[email protected]>

[1]: https://lkml.org/lkml/2016/11/4/245
[2]: https://lkml.org/lkml/2016/11/4/253
---
v1: https://lkml.org/lkml/2017/5/2/404

---
 drivers/ata/libata-sff.c | 44 ++++++++------------------------------------
 1 file changed, 8 insertions(+), 36 deletions(-)
drivers/ata/libata-sff.c +8 -36
index 2bd92dca3e62..01cf07c919bc 100644
@@ -716,24 +716,10 @@ static void ata_pio_sector(struct ata_queued_cmd *qc)
 
 	DPRINTK("data %s\n", qc->tf.flags & ATA_TFLAG_WRITE ? "write" : "read");
 
-	if (PageHighMem(page)) {
-		unsigned long flags;
-
-		/* FIXME: use a bounce buffer */
-		local_irq_save(flags);
-		buf = kmap_atomic(page);
-
-		/* do the actual data transfer */
-		ap->ops->sff_data_xfer(qc, buf + offset, qc->sect_size,
-				       do_write);
-
-		kunmap_atomic(buf);
-		local_irq_restore(flags);
-	} else {
-		buf = page_address(page);
-		ap->ops->sff_data_xfer(qc, buf + offset, qc->sect_size,
-				       do_write);
-	}
+	/* do the actual data transfer */
+	buf = kmap_atomic(page);
+	ap->ops->sff_data_xfer(qc, buf + offset, qc->sect_size, do_write);
+	kunmap_atomic(buf);
 
 	if (!do_write && !PageSlab(page))
 		flush_dcache_page(page);

@@ -861,24 +847,10 @@ static int __atapi_pio_bytes(struct ata_queued_cmd *qc, unsigned int bytes)
 
 	DPRINTK("data %s\n", qc->tf.flags & ATA_TFLAG_WRITE ? "write" : "read");
 
-	if (PageHighMem(page)) {
-		unsigned long flags;
-
-		/* FIXME: use bounce buffer */
-		local_irq_save(flags);
-		buf = kmap_atomic(page);
-
-		/* do the actual data transfer */
-		consumed = ap->ops->sff_data_xfer(qc, buf + offset,
-								count, rw);
-
-		kunmap_atomic(buf);
-		local_irq_restore(flags);
-	} else {
-		buf = page_address(page);
-		consumed = ap->ops->sff_data_xfer(qc, buf + offset,
-								count, rw);
-	}
+	/* do the actual data transfer */
+	buf = kmap_atomic(page);
+	consumed = ap->ops->sff_data_xfer(qc, buf + offset, count, rw);
+	kunmap_atomic(buf);
 
 	bytes -= min(bytes, consumed);
 	qc->curbytes += count;
Signature · 1 line
2.11.0
Applied
#2 lore raw
In reply to Tycho Andersen
Quoted 9 lines · Tycho Andersen

The XPFO [1] patchset may unmap pages from physmap if they happened to be destined for userspace. If such a page is unmapped, it needs to be remapped. Rather than test if a page is in the highmem/xpfo unmapped state, Christoph suggested [2] that we simply always map the page.

v2: * drop comment about bounce buffer
* don't save IRQs before kmap/unmap
* formatting

Quoted 7 lines
Suggested-by Christoph Hellwig <[email protected]>
Signed-off-by Tycho Andersen <[email protected]>
CC Juerg Haefliger <[email protected]>
CC Tejun Heo <[email protected]>

[1]: https://lkml.org/lkml/2016/11/4/245
[2]: https://lkml.org/lkml/2016/11/4/253

Yeah, it's a nice cleanup.

Applied to libata/for-4.13.

Thanks.

Signature · 1 line
tejun