| Message ID | 20240731191312.1710417-1-robh@kernel.org |
|---|---|
| State | New |
| Headers | show |
| Series | bus: ti-sysc: Use of_property_present() | expand |
On 31/07/2024 22:12, Rob Herring (Arm) wrote: > Use of_property_present() to test for property presence rather than > of_get_property(). This is part of a larger effort to remove callers > of of_get_property() and similar functions. of_get_property() leaks > the DT property data pointer which is a problem for dynamically > allocated nodes which may be freed. > > The code was also incorrectly assigning the return value to a 'struct > property' pointer. It didn't matter as "prop" was never dereferenced. > > Signed-off-by: Rob Herring (Arm) <robh@kernel.org> Reviewed by: Roger Quadros <rogerq@kernel.org>
Hi Roger, Roger Quadros <rogerq@kernel.org> writes: > On 31/07/2024 22:12, Rob Herring (Arm) wrote: >> Use of_property_present() to test for property presence rather than >> of_get_property(). This is part of a larger effort to remove callers >> of of_get_property() and similar functions. of_get_property() leaks >> the DT property data pointer which is a problem for dynamically >> allocated nodes which may be freed. >> >> The code was also incorrectly assigning the return value to a 'struct >> property' pointer. It didn't matter as "prop" was never dereferenced. >> >> Signed-off-by: Rob Herring (Arm) <robh@kernel.org> > > Reviewed by: Roger Quadros <rogerq@kernel.org> just FYI, missing a '-' in your Reviewed-by, which means tools like b4 will not spot it. I added it manually this time cuz I happened to notice it was missing. Kevin
On Wed, 31 Jul 2024 13:12:39 -0600, Rob Herring (Arm) wrote: > Use of_property_present() to test for property presence rather than > of_get_property(). This is part of a larger effort to remove callers > of of_get_property() and similar functions. of_get_property() leaks > the DT property data pointer which is a problem for dynamically > allocated nodes which may be freed. > > The code was also incorrectly assigning the return value to a 'struct > property' pointer. It didn't matter as "prop" was never dereferenced. > > [...] Applied, thanks! [1/1] bus: ti-sysc: Use of_property_present() commit: 0070dc29c85f0859a6071844b88fca6bce2974e4 Best regards,
Hi Kevin, On 05/08/2024 20:36, Kevin Hilman wrote: > Hi Roger, > > Roger Quadros <rogerq@kernel.org> writes: > >> On 31/07/2024 22:12, Rob Herring (Arm) wrote: >>> Use of_property_present() to test for property presence rather than >>> of_get_property(). This is part of a larger effort to remove callers >>> of of_get_property() and similar functions. of_get_property() leaks >>> the DT property data pointer which is a problem for dynamically >>> allocated nodes which may be freed. >>> >>> The code was also incorrectly assigning the return value to a 'struct >>> property' pointer. It didn't matter as "prop" was never dereferenced. >>> >>> Signed-off-by: Rob Herring (Arm) <robh@kernel.org> >> >> Reviewed by: Roger Quadros <rogerq@kernel.org> > > just FYI, missing a '-' in your Reviewed-by, which means tools like b4 > will not spot it. I added it manually this time cuz I happened to > notice it was missing. Sorry, my bad. Thanks for fixing this up.
diff --git a/drivers/bus/ti-sysc.c b/drivers/bus/ti-sysc.c index 2b59ef61dda2..7caaf89f3bf2 100644 --- a/drivers/bus/ti-sysc.c +++ b/drivers/bus/ti-sysc.c @@ -2569,14 +2569,12 @@ static const struct sysc_dts_quirk sysc_dts_quirks[] = { static void sysc_parse_dts_quirks(struct sysc *ddata, struct device_node *np, bool is_child) { - const struct property *prop; - int i, len; + int i; for (i = 0; i < ARRAY_SIZE(sysc_dts_quirks); i++) { const char *name = sysc_dts_quirks[i].name; - prop = of_get_property(np, name, &len); - if (!prop) + if (!of_property_present(np, name)) continue; ddata->cfg.quirks |= sysc_dts_quirks[i].mask;
Use of_property_present() to test for property presence rather than of_get_property(). This is part of a larger effort to remove callers of of_get_property() and similar functions. of_get_property() leaks the DT property data pointer which is a problem for dynamically allocated nodes which may be freed. The code was also incorrectly assigning the return value to a 'struct property' pointer. It didn't matter as "prop" was never dereferenced. Signed-off-by: Rob Herring (Arm) <robh@kernel.org> --- drivers/bus/ti-sysc.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-)