2022-08-28 20:58:00

by Li Zhong

Subject: [PATCH v1] drivers/ata/libata-core: check return value of sata_scr_read()

sata_scr_read() could return 0 on failure. Check the return value.

Signed-off-by: Li Zhong <[email protected]>
---
drivers/ata/libata-core.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 826d41f341e4..ae08c7d35cb0 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -3021,7 +3021,8 @@ static void sata_print_link_status(struct ata_link *link)

if (sata_scr_read(link, SCR_STATUS, &sstatus))
return;
- sata_scr_read(link, SCR_CONTROL, &scontrol);
+ if (sata_scr_read(link, SCR_CONTROL, &scontrol))
+ return;

if (ata_phys_link_online(link)) {
tmp = (sstatus >> 4) & 0xf;
--
2.25.1

2022-08-30 01:46:52

by Damien Le Moal

Subject: Re: [PATCH v1] drivers/ata/libata-core: check return value of sata_scr_read()

On 2022/08/29 5:45, Li Zhong wrote:
> sata_scr_read() could return 0 on failure. Check the return value.

Looks OK as a fix but the above sentence has a bug. 0 indicates success, not
failure. So this should read:

sata_scr_read() may fail with a negative error code. Check that function return
value when reading the control register.

Also, the patch title should be something like:

ata: libata-core: Check errors in sata_print_link_status()

>
> Signed-off-by: Li Zhong <[email protected]>
> ---
> drivers/ata/libata-core.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
> index 826d41f341e4..ae08c7d35cb0 100644
> --- a/drivers/ata/libata-core.c
> +++ b/drivers/ata/libata-core.c
> @@ -3021,7 +3021,8 @@ static void sata_print_link_status(struct ata_link *link)
>
> if (sata_scr_read(link, SCR_STATUS, &sstatus))
> return;
> - sata_scr_read(link, SCR_CONTROL, &scontrol);
> + if (sata_scr_read(link, SCR_CONTROL, &scontrol))
> + return;
>
> if (ata_phys_link_online(link)) {
> tmp = (sstatus >> 4) & 0xf;

--
Damien Le Moal
Western Digital Research

2022-09-03 23:29:43

by Li Zhong

Subject: Re: [PATCH v1] drivers/ata/libata-core: check return value of sata_scr_read()

On Mon, Aug 29, 2022 at 6:38 PM Damien Le Moal <[email protected]> wrote:
>
> On 2022/08/29 5:45, Li Zhong wrote:
> > sata_scr_read() could return 0 on failure. Check the return value.
>
> Looks OK as a fix but the above sentence has a bug. 0 indicates success, not
> failure. So this should read:
>
> sata_scr_read() may fail with a negative error code. Check that function return
> value when reading the control register.
>
> Also, the patch title should be something like:
>
> ata: libata-core: Check errors in sata_print_link_status()
>

Thanks. Will fix it in the new version.

> >
> > Signed-off-by: Li Zhong <[email protected]>
> > ---
> > drivers/ata/libata-core.c | 3 ++-
> > 1 file changed, 2 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
> > index 826d41f341e4..ae08c7d35cb0 100644
> > --- a/drivers/ata/libata-core.c
> > +++ b/drivers/ata/libata-core.c
> > @@ -3021,7 +3021,8 @@ static void sata_print_link_status(struct ata_link *link)
> >
> > if (sata_scr_read(link, SCR_STATUS, &sstatus))
> > return;
> > - sata_scr_read(link, SCR_CONTROL, &scontrol);
> > + if (sata_scr_read(link, SCR_CONTROL, &scontrol))
> > + return;
> >
> > if (ata_phys_link_online(link)) {
> > tmp = (sstatus >> 4) & 0xf;
>
> --
> Damien Le Moal
> Western Digital Research
>