| Message ID | 20150212095323.1829.45732.stgit@localhost.localdomain (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
On Thu, 2015-02-12 at 15:23 +0530, Vasant Hegde wrote: > Present code checks for update_flash_data in opal_flash_term_callback(). > update_flash_data has been statically initialized to zero, and that > is the value of FLASH_IMG_READY. Also code update initialization happens > during subsys init. Please statically initialize it to a sane value instead. > So if reboot is issued before the subsys init stage then we endup displaying > "Flashing new firmware" message.. which may confuse end user. > > This patch adds additional validation to make sure image is actually loaded > or not. > > Reported-by: Sam Bobroff <sam.bobroff@au1.ibm.com> > Signed-off-by: Vasant Hegde <hegdevasant@linux.vnet.ibm.com> > --- > arch/powerpc/platforms/powernv/opal-flash.c | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) > > diff --git a/arch/powerpc/platforms/powernv/opal-flash.c b/arch/powerpc/platforms/powernv/opal-flash.c > index 5c21d9c..5455cd4 100644 > --- a/arch/powerpc/platforms/powernv/opal-flash.c > +++ b/arch/powerpc/platforms/powernv/opal-flash.c > @@ -319,7 +319,8 @@ void opal_flash_term_callback(void) > { > struct cpumask mask; > > - if (update_flash_data.status != FLASH_IMG_READY) > + if (update_flash_data.status != FLASH_IMG_READY || > + image_data.status != IMAGE_READY) > return; > > pr_alert("FLASH: Flashing new firmware\n");
On 13/02/15 08:27, Benjamin Herrenschmidt wrote: > On Thu, 2015-02-12 at 15:23 +0530, Vasant Hegde wrote: >> Present code checks for update_flash_data in opal_flash_term_callback(). >> update_flash_data has been statically initialized to zero, and that >> is the value of FLASH_IMG_READY. Also code update initialization happens >> during subsys init. > > Please statically initialize it to a sane value instead. I've tested this patch and it works for me (the message is suppressed) but I agree with Ben that it seems cleaner to use a static initializer. >> So if reboot is issued before the subsys init stage then we endup displaying >> "Flashing new firmware" message.. which may confuse end user. >> >> This patch adds additional validation to make sure image is actually loaded >> or not. >> >> Reported-by: Sam Bobroff <sam.bobroff@au1.ibm.com> >> Signed-off-by: Vasant Hegde <hegdevasant@linux.vnet.ibm.com> >> --- >> arch/powerpc/platforms/powernv/opal-flash.c | 3 ++- >> 1 file changed, 2 insertions(+), 1 deletion(-) >> >> diff --git a/arch/powerpc/platforms/powernv/opal-flash.c b/arch/powerpc/platforms/powernv/opal-flash.c >> index 5c21d9c..5455cd4 100644 >> --- a/arch/powerpc/platforms/powernv/opal-flash.c >> +++ b/arch/powerpc/platforms/powernv/opal-flash.c >> @@ -319,7 +319,8 @@ void opal_flash_term_callback(void) >> { >> struct cpumask mask; >> >> - if (update_flash_data.status != FLASH_IMG_READY) >> + if (update_flash_data.status != FLASH_IMG_READY || >> + image_data.status != IMAGE_READY) >> return; >> >> pr_alert("FLASH: Flashing new firmware\n"); > > > _______________________________________________ > Linuxppc-dev mailing list > Linuxppc-dev@lists.ozlabs.org > https://lists.ozlabs.org/listinfo/linuxppc-dev >
On 02/13/2015 02:57 AM, Benjamin Herrenschmidt wrote: > On Thu, 2015-02-12 at 15:23 +0530, Vasant Hegde wrote: >> Present code checks for update_flash_data in opal_flash_term_callback(). >> update_flash_data has been statically initialized to zero, and that >> is the value of FLASH_IMG_READY. Also code update initialization happens >> during subsys init. > > Please statically initialize it to a sane value instead. Ok. Will fix and send v2 soon. Thanks! -Vasant > >> So if reboot is issued before the subsys init stage then we endup displaying >> "Flashing new firmware" message.. which may confuse end user. >> >> This patch adds additional validation to make sure image is actually loaded >> or not. >> >> Reported-by: Sam Bobroff <sam.bobroff@au1.ibm.com> >> Signed-off-by: Vasant Hegde <hegdevasant@linux.vnet.ibm.com> >> --- >> arch/powerpc/platforms/powernv/opal-flash.c | 3 ++- >> 1 file changed, 2 insertions(+), 1 deletion(-) >> >> diff --git a/arch/powerpc/platforms/powernv/opal-flash.c b/arch/powerpc/platforms/powernv/opal-flash.c >> index 5c21d9c..5455cd4 100644 >> --- a/arch/powerpc/platforms/powernv/opal-flash.c >> +++ b/arch/powerpc/platforms/powernv/opal-flash.c >> @@ -319,7 +319,8 @@ void opal_flash_term_callback(void) >> { >> struct cpumask mask; >> >> - if (update_flash_data.status != FLASH_IMG_READY) >> + if (update_flash_data.status != FLASH_IMG_READY || >> + image_data.status != IMAGE_READY) >> return; >> >> pr_alert("FLASH: Flashing new firmware\n"); > >
On 02/13/2015 08:16 AM, Sam Bobroff wrote: > On 13/02/15 08:27, Benjamin Herrenschmidt wrote: >> On Thu, 2015-02-12 at 15:23 +0530, Vasant Hegde wrote: >>> Present code checks for update_flash_data in opal_flash_term_callback(). >>> update_flash_data has been statically initialized to zero, and that >>> is the value of FLASH_IMG_READY. Also code update initialization happens >>> during subsys init. >> >> Please statically initialize it to a sane value instead. > > I've tested this patch and it works for me (the message is suppressed) Sam, Thanks for testing! -Vasant > but I agree with Ben that it seems cleaner to use a static initializer. > >>> So if reboot is issued before the subsys init stage then we endup displaying >>> "Flashing new firmware" message.. which may confuse end user. >>> >>> This patch adds additional validation to make sure image is actually loaded >>> or not. >>> >>> Reported-by: Sam Bobroff <sam.bobroff@au1.ibm.com> >>> Signed-off-by: Vasant Hegde <hegdevasant@linux.vnet.ibm.com> >>> --- >>> arch/powerpc/platforms/powernv/opal-flash.c | 3 ++- >>> 1 file changed, 2 insertions(+), 1 deletion(-) >>> >>> diff --git a/arch/powerpc/platforms/powernv/opal-flash.c b/arch/powerpc/platforms/powernv/opal-flash.c >>> index 5c21d9c..5455cd4 100644 >>> --- a/arch/powerpc/platforms/powernv/opal-flash.c >>> +++ b/arch/powerpc/platforms/powernv/opal-flash.c >>> @@ -319,7 +319,8 @@ void opal_flash_term_callback(void) >>> { >>> struct cpumask mask; >>> >>> - if (update_flash_data.status != FLASH_IMG_READY) >>> + if (update_flash_data.status != FLASH_IMG_READY || >>> + image_data.status != IMAGE_READY) >>> return; >>> >>> pr_alert("FLASH: Flashing new firmware\n"); >> >> >> _______________________________________________ >> Linuxppc-dev mailing list >> Linuxppc-dev@lists.ozlabs.org >> https://lists.ozlabs.org/listinfo/linuxppc-dev >> > > _______________________________________________ > Linuxppc-dev mailing list > Linuxppc-dev@lists.ozlabs.org > https://lists.ozlabs.org/listinfo/linuxppc-dev >
diff --git a/arch/powerpc/platforms/powernv/opal-flash.c b/arch/powerpc/platforms/powernv/opal-flash.c index 5c21d9c..5455cd4 100644 --- a/arch/powerpc/platforms/powernv/opal-flash.c +++ b/arch/powerpc/platforms/powernv/opal-flash.c @@ -319,7 +319,8 @@ void opal_flash_term_callback(void) { struct cpumask mask; - if (update_flash_data.status != FLASH_IMG_READY) + if (update_flash_data.status != FLASH_IMG_READY || + image_data.status != IMAGE_READY) return; pr_alert("FLASH: Flashing new firmware\n");
Present code checks for update_flash_data in opal_flash_term_callback(). update_flash_data has been statically initialized to zero, and that is the value of FLASH_IMG_READY. Also code update initialization happens during subsys init. So if reboot is issued before the subsys init stage then we endup displaying "Flashing new firmware" message.. which may confuse end user. This patch adds additional validation to make sure image is actually loaded or not. Reported-by: Sam Bobroff <sam.bobroff@au1.ibm.com> Signed-off-by: Vasant Hegde <hegdevasant@linux.vnet.ibm.com> --- arch/powerpc/platforms/powernv/opal-flash.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)