| Message ID | 20180109094702.21488-1-s.hauer@pengutronix.de |
|---|---|
| State | Accepted |
| Delegated to: | Boris Brezillon |
| Headers | show |
| Series | [v2] mtd: tests: nandbiterrs: Fix read_page return value | expand |
On Tue, 9 Jan 2018 10:47:02 +0100 Sascha Hauer <s.hauer@pengutronix.de> wrote: > The number of corrected bitflips is not correctly reported by > the test until the bitflip threshold is reached. > > read_page() shall return the number of corrected bitflips, but > mtd_read() returns 0 or a negative error, so we can't forward > its return value. In the absence of an error we always have > calculate the number of bitflips ourselves. > > Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de> I'll apply this patch, but you should really start using the userspace tests provided by mtd-utils, because I still plan to remove the in-kernel test modules at some point ;-). > --- > drivers/mtd/tests/nandbiterrs.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/drivers/mtd/tests/nandbiterrs.c b/drivers/mtd/tests/nandbiterrs.c > index 5f03b8c885a9..cde19c99e77b 100644 > --- a/drivers/mtd/tests/nandbiterrs.c > +++ b/drivers/mtd/tests/nandbiterrs.c > @@ -151,7 +151,7 @@ static int read_page(int log) > memcpy(&oldstats, &mtd->ecc_stats, sizeof(oldstats)); > > err = mtd_read(mtd, offset, mtd->writesize, &read, rbuffer); > - if (err == -EUCLEAN) > + if (!err || err == -EUCLEAN) > err = mtd->ecc_stats.corrected - oldstats.corrected; > > if (err < 0 || read != mtd->writesize) {
On Tue, Jan 09, 2018 at 10:51:08AM +0100, Boris Brezillon wrote: > On Tue, 9 Jan 2018 10:47:02 +0100 > Sascha Hauer <s.hauer@pengutronix.de> wrote: > > > The number of corrected bitflips is not correctly reported by > > the test until the bitflip threshold is reached. > > > > read_page() shall return the number of corrected bitflips, but > > mtd_read() returns 0 or a negative error, so we can't forward > > its return value. In the absence of an error we always have > > calculate the number of bitflips ourselves. > > > > Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de> > > I'll apply this patch, but you should really start using the userspace > tests provided by mtd-utils, because I still plan to remove the > in-kernel test modules at some point ;-). Please do so. When I cannot find the in-kernel tests anymore I'll recall that there are tests in userspace ;) Sascha
On Tue, 9 Jan 2018 10:47:02 +0100 Sascha Hauer <s.hauer@pengutronix.de> wrote: > The number of corrected bitflips is not correctly reported by > the test until the bitflip threshold is reached. > > read_page() shall return the number of corrected bitflips, but > mtd_read() returns 0 or a negative error, so we can't forward > its return value. In the absence of an error we always have > calculate the number of bitflips ourselves. > Applied. Thanks, Boris > Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de> > --- > drivers/mtd/tests/nandbiterrs.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/drivers/mtd/tests/nandbiterrs.c b/drivers/mtd/tests/nandbiterrs.c > index 5f03b8c885a9..cde19c99e77b 100644 > --- a/drivers/mtd/tests/nandbiterrs.c > +++ b/drivers/mtd/tests/nandbiterrs.c > @@ -151,7 +151,7 @@ static int read_page(int log) > memcpy(&oldstats, &mtd->ecc_stats, sizeof(oldstats)); > > err = mtd_read(mtd, offset, mtd->writesize, &read, rbuffer); > - if (err == -EUCLEAN) > + if (!err || err == -EUCLEAN) > err = mtd->ecc_stats.corrected - oldstats.corrected; > > if (err < 0 || read != mtd->writesize) {
diff --git a/drivers/mtd/tests/nandbiterrs.c b/drivers/mtd/tests/nandbiterrs.c index 5f03b8c885a9..cde19c99e77b 100644 --- a/drivers/mtd/tests/nandbiterrs.c +++ b/drivers/mtd/tests/nandbiterrs.c @@ -151,7 +151,7 @@ static int read_page(int log) memcpy(&oldstats, &mtd->ecc_stats, sizeof(oldstats)); err = mtd_read(mtd, offset, mtd->writesize, &read, rbuffer); - if (err == -EUCLEAN) + if (!err || err == -EUCLEAN) err = mtd->ecc_stats.corrected - oldstats.corrected; if (err < 0 || read != mtd->writesize) {
The number of corrected bitflips is not correctly reported by the test until the bitflip threshold is reached. read_page() shall return the number of corrected bitflips, but mtd_read() returns 0 or a negative error, so we can't forward its return value. In the absence of an error we always have calculate the number of bitflips ourselves. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de> --- drivers/mtd/tests/nandbiterrs.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)