[<prev] [next>] [day] [month] [year] [list]
Message-ID: <148192600397.40331.16038852518288391005.stgit@dwillia2-desk3.amr.corp.intel.com>
Date: Fri, 16 Dec 2016 14:06:44 -0800
From: Dan Williams <dan.j.williams@...el.com>
To: linux-nvdimm@...ts.01.org
Cc: Fabian Frederick <fabf@...net.be>,
Colin Ian King <colin.king@...onical.com>,
Dave Jiang <dave.jiang@...el.com>, linux-kernel@...r.kernel.org
Subject: [PATCH] libnvdimm: fix mishandled nvdimm_clear_poison() return value
Colin, via static analysis, reports that the length could be negative
from nvdimm_clear_poison() in the error case. There was a similar
problem with commit 0a3f27b9a6a8 "libnvdimm, namespace: avoid multiple
sector calculations" that I noticed when merging the for-4.10/libnvdimm
topic branch into libnvdimm-for-next, but I missed this one. Fix both of
them to the following procedure:
* if we clear a block's worth of media, clear that many blocks in
badblocks
* if we clear less than the requested size of the transfer return an
error
* always invalidate cache after any non-error / non-zero
nvdimm_clear_poison result
Fixes: 82bf1037f2ca ("libnvdimm: check and clear poison before writing to pmem")
Fixes: 0a3f27b9a6a8 ("libnvdimm, namespace: avoid multiple sector calculations")
Cc: Fabian Frederick <fabf@...net.be>
Cc: Dave Jiang <dave.jiang@...el.com>
Reported-by: Colin Ian King <colin.king@...onical.com>
Signed-off-by: Dan Williams <dan.j.williams@...el.com>
---
drivers/nvdimm/claim.c | 9 +++++----
drivers/nvdimm/pmem.c | 21 ++++++++++++++-------
2 files changed, 19 insertions(+), 11 deletions(-)
diff --git a/drivers/nvdimm/claim.c b/drivers/nvdimm/claim.c
index 97d1772774a8..b3323c0697f6 100644
--- a/drivers/nvdimm/claim.c
+++ b/drivers/nvdimm/claim.c
@@ -247,12 +247,13 @@ static int nsio_rw_bytes(struct nd_namespace_common *ndns,
long cleared;
cleared = nvdimm_clear_poison(&ndns->dev, offset, size);
- if (cleared != size) {
- size = cleared;
+ if (cleared < size)
rc = -EIO;
+ if (cleared > 0 && cleared / 512) {
+ cleared /= 512;
+ badblocks_clear(&nsio->bb, sector, cleared);
}
-
- badblocks_clear(&nsio->bb, sector, cleared >> 9);
+ invalidate_pmem(nsio->addr + offset, size);
} else
rc = -EIO;
}
diff --git a/drivers/nvdimm/pmem.c b/drivers/nvdimm/pmem.c
index bcc359a4e64d..ecf79fd64517 100644
--- a/drivers/nvdimm/pmem.c
+++ b/drivers/nvdimm/pmem.c
@@ -47,22 +47,29 @@ static struct nd_region *to_region(struct pmem_device *pmem)
return to_nd_region(to_dev(pmem)->parent);
}
-static void pmem_clear_poison(struct pmem_device *pmem, phys_addr_t offset,
+static int pmem_clear_poison(struct pmem_device *pmem, phys_addr_t offset,
unsigned int len)
{
struct device *dev = to_dev(pmem);
- sector_t sector, cleared;
+ sector_t sector;
+ long cleared;
+ int rc = 0;
sector = (offset - pmem->data_offset) / 512;
- cleared = nvdimm_clear_poison(dev, pmem->phys_addr + offset, len) / 512;
- if (cleared) {
- dev_dbg(dev, "%s: %#llx clear %ld sector%s\n",
- __func__, (unsigned long long) sector,
- cleared, cleared > 1 ? "s" : "");
+ cleared = nvdimm_clear_poison(dev, pmem->phys_addr + offset, len);
+ if (cleared < len)
+ rc = -EIO;
+ if (cleared > 0 && cleared / 512) {
+ cleared /= 512;
+ dev_dbg(dev, "%s: %#llx clear %ld sector%s\n", __func__,
+ (unsigned long long) sector, cleared,
+ cleared > 1 ? "s" : "");
badblocks_clear(&pmem->bb, sector, cleared);
}
invalidate_pmem(pmem->virt_addr + offset, len);
+
+ return rc;
}
static void write_pmem(void *pmem_addr, struct page *page,
Powered by blists - more mailing lists