Message ID | 20190615051722.13994-1-rfried.dev@gmail.com (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
Series | net: cadence_gem: clear RX control descriptor | expand |
ping On Sat, Jun 15, 2019 at 8:17 AM Ramon Fried <rfried.dev@gmail.com> wrote: > > The RX ring descriptors control field is used for setting > SOF and EOF (start of frame and end of frame). > The SOF and EOF weren't cleared from the previous descriptors, > causing inconsistencies in ring buffer. > Fix that by clearing the control field of every descriptors we're > processing. > > Signed-off-by: Ramon Fried <rfried.dev@gmail.com> > --- > hw/net/cadence_gem.c | 7 +++++++ > 1 file changed, 7 insertions(+) > > diff --git a/hw/net/cadence_gem.c b/hw/net/cadence_gem.c > index ecee22525c..d83a82bdb0 100644 > --- a/hw/net/cadence_gem.c > +++ b/hw/net/cadence_gem.c > @@ -406,6 +406,11 @@ static inline void rx_desc_set_sof(uint32_t *desc) > desc[1] |= DESC_1_RX_SOF; > } > > +static inline void rx_desc_clear(uint32_t *desc) > +{ > + desc[1] = 0; > +} > + > static inline void rx_desc_set_eof(uint32_t *desc) > { > desc[1] |= DESC_1_RX_EOF; > @@ -994,6 +999,8 @@ static ssize_t gem_receive(NetClientState *nc, const uint8_t *buf, size_t size) > bytes_to_copy -= MIN(bytes_to_copy, rxbufsize); > > /* Update the descriptor. */ > + rx_desc_clear(s->rx_desc[q]); > + > if (first_desc) { > rx_desc_set_sof(s->rx_desc[q]); > first_desc = false; > -- > 2.21.0 >
On 7/16/19 9:11 AM, Ramon Fried wrote: > ping > > On Sat, Jun 15, 2019 at 8:17 AM Ramon Fried <rfried.dev@gmail.com> wrote: >> >> The RX ring descriptors control field is used for setting >> SOF and EOF (start of frame and end of frame). >> The SOF and EOF weren't cleared from the previous descriptors, >> causing inconsistencies in ring buffer. >> Fix that by clearing the control field of every descriptors we're >> processing. >> >> Signed-off-by: Ramon Fried <rfried.dev@gmail.com> >> --- >> hw/net/cadence_gem.c | 7 +++++++ >> 1 file changed, 7 insertions(+) >> >> diff --git a/hw/net/cadence_gem.c b/hw/net/cadence_gem.c >> index ecee22525c..d83a82bdb0 100644 >> --- a/hw/net/cadence_gem.c >> +++ b/hw/net/cadence_gem.c >> @@ -406,6 +406,11 @@ static inline void rx_desc_set_sof(uint32_t *desc) >> desc[1] |= DESC_1_RX_SOF; >> } >> >> +static inline void rx_desc_clear(uint32_t *desc) >> +{ >> + desc[1] = 0; >> +} >> + >> static inline void rx_desc_set_eof(uint32_t *desc) >> { >> desc[1] |= DESC_1_RX_EOF; >> @@ -994,6 +999,8 @@ static ssize_t gem_receive(NetClientState *nc, const uint8_t *buf, size_t size) >> bytes_to_copy -= MIN(bytes_to_copy, rxbufsize); >> >> /* Update the descriptor. */ >> + rx_desc_clear(s->rx_desc[q]); Maybe move the call before the comment, regardless: Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> >> + >> if (first_desc) { >> rx_desc_set_sof(s->rx_desc[q]); >> first_desc = false; >> -- >> 2.21.0 >>
On 2019/6/15 下午1:17, Ramon Fried wrote: > The RX ring descriptors control field is used for setting > SOF and EOF (start of frame and end of frame). > The SOF and EOF weren't cleared from the previous descriptors, > causing inconsistencies in ring buffer. > Fix that by clearing the control field of every descriptors we're > processing. > > Signed-off-by: Ramon Fried <rfried.dev@gmail.com> > --- > hw/net/cadence_gem.c | 7 +++++++ > 1 file changed, 7 insertions(+) > > diff --git a/hw/net/cadence_gem.c b/hw/net/cadence_gem.c > index ecee22525c..d83a82bdb0 100644 > --- a/hw/net/cadence_gem.c > +++ b/hw/net/cadence_gem.c > @@ -406,6 +406,11 @@ static inline void rx_desc_set_sof(uint32_t *desc) > desc[1] |= DESC_1_RX_SOF; > } > > +static inline void rx_desc_clear(uint32_t *desc) Nit: is this better to name this as "rx_desc_clear_control()" ? Thanks > +{ > + desc[1] = 0; > +} > + > static inline void rx_desc_set_eof(uint32_t *desc) > { > desc[1] |= DESC_1_RX_EOF; > @@ -994,6 +999,8 @@ static ssize_t gem_receive(NetClientState *nc, const uint8_t *buf, size_t size) > bytes_to_copy -= MIN(bytes_to_copy, rxbufsize); > > /* Update the descriptor. */ > + rx_desc_clear(s->rx_desc[q]); > + > if (first_desc) { > rx_desc_set_sof(s->rx_desc[q]); > first_desc = false;
On Tue, Jul 16, 2019 at 10:19 AM Philippe Mathieu-Daudé <philmd@redhat.com> wrote: > > On 7/16/19 9:11 AM, Ramon Fried wrote: > > ping > > > > On Sat, Jun 15, 2019 at 8:17 AM Ramon Fried <rfried.dev@gmail.com> wrote: > >> > >> The RX ring descriptors control field is used for setting > >> SOF and EOF (start of frame and end of frame). > >> The SOF and EOF weren't cleared from the previous descriptors, > >> causing inconsistencies in ring buffer. > >> Fix that by clearing the control field of every descriptors we're > >> processing. > >> > >> Signed-off-by: Ramon Fried <rfried.dev@gmail.com> > >> --- > >> hw/net/cadence_gem.c | 7 +++++++ > >> 1 file changed, 7 insertions(+) > >> > >> diff --git a/hw/net/cadence_gem.c b/hw/net/cadence_gem.c > >> index ecee22525c..d83a82bdb0 100644 > >> --- a/hw/net/cadence_gem.c > >> +++ b/hw/net/cadence_gem.c > >> @@ -406,6 +406,11 @@ static inline void rx_desc_set_sof(uint32_t *desc) > >> desc[1] |= DESC_1_RX_SOF; > >> } > >> > >> +static inline void rx_desc_clear(uint32_t *desc) > >> +{ > >> + desc[1] = 0; > >> +} > >> + > >> static inline void rx_desc_set_eof(uint32_t *desc) > >> { > >> desc[1] |= DESC_1_RX_EOF; > >> @@ -994,6 +999,8 @@ static ssize_t gem_receive(NetClientState *nc, const uint8_t *buf, size_t size) > >> bytes_to_copy -= MIN(bytes_to_copy, rxbufsize); > >> > >> /* Update the descriptor. */ > >> + rx_desc_clear(s->rx_desc[q]); > > Maybe move the call before the comment, regardless: Make sense, sent v2. > Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> > > >> + > >> if (first_desc) { > >> rx_desc_set_sof(s->rx_desc[q]); > >> first_desc = false; > >> -- > >> 2.21.0 > >>
On Tue, Jul 16, 2019 at 11:42 AM Jason Wang <jasowang@redhat.com> wrote: > > > On 2019/6/15 下午1:17, Ramon Fried wrote: > > The RX ring descriptors control field is used for setting > > SOF and EOF (start of frame and end of frame). > > The SOF and EOF weren't cleared from the previous descriptors, > > causing inconsistencies in ring buffer. > > Fix that by clearing the control field of every descriptors we're > > processing. > > > > Signed-off-by: Ramon Fried <rfried.dev@gmail.com> > > --- > > hw/net/cadence_gem.c | 7 +++++++ > > 1 file changed, 7 insertions(+) > > > > diff --git a/hw/net/cadence_gem.c b/hw/net/cadence_gem.c > > index ecee22525c..d83a82bdb0 100644 > > --- a/hw/net/cadence_gem.c > > +++ b/hw/net/cadence_gem.c > > @@ -406,6 +406,11 @@ static inline void rx_desc_set_sof(uint32_t *desc) > > desc[1] |= DESC_1_RX_SOF; > > } > > > > +static inline void rx_desc_clear(uint32_t *desc) > > > Nit: is this better to name this as "rx_desc_clear_control()" ? I agree, sent v2. > > Thanks > > > > +{ > > + desc[1] = 0; > > +} > > + > > static inline void rx_desc_set_eof(uint32_t *desc) > > { > > desc[1] |= DESC_1_RX_EOF; > > @@ -994,6 +999,8 @@ static ssize_t gem_receive(NetClientState *nc, const uint8_t *buf, size_t size) > > bytes_to_copy -= MIN(bytes_to_copy, rxbufsize); > > > > /* Update the descriptor. */ > > + rx_desc_clear(s->rx_desc[q]); > > + > > if (first_desc) { > > rx_desc_set_sof(s->rx_desc[q]); > > first_desc = false; > > >
diff --git a/hw/net/cadence_gem.c b/hw/net/cadence_gem.c index ecee22525c..d83a82bdb0 100644 --- a/hw/net/cadence_gem.c +++ b/hw/net/cadence_gem.c @@ -406,6 +406,11 @@ static inline void rx_desc_set_sof(uint32_t *desc) desc[1] |= DESC_1_RX_SOF; } +static inline void rx_desc_clear(uint32_t *desc) +{ + desc[1] = 0; +} + static inline void rx_desc_set_eof(uint32_t *desc) { desc[1] |= DESC_1_RX_EOF; @@ -994,6 +999,8 @@ static ssize_t gem_receive(NetClientState *nc, const uint8_t *buf, size_t size) bytes_to_copy -= MIN(bytes_to_copy, rxbufsize); /* Update the descriptor. */ + rx_desc_clear(s->rx_desc[q]); + if (first_desc) { rx_desc_set_sof(s->rx_desc[q]); first_desc = false;
The RX ring descriptors control field is used for setting SOF and EOF (start of frame and end of frame). The SOF and EOF weren't cleared from the previous descriptors, causing inconsistencies in ring buffer. Fix that by clearing the control field of every descriptors we're processing. Signed-off-by: Ramon Fried <rfried.dev@gmail.com> --- hw/net/cadence_gem.c | 7 +++++++ 1 file changed, 7 insertions(+)