Hi all,
Today's linux-next merge of the v4l-dvb tree got a conflict in:
drivers/media/i2c/ds90ub960.c
between commits:
3ec29d51b546 ("media: i2c: ds90ub960: Protect alias_use_mask with a mutex")
818bd489f137 ("i2c: use client addresses directly in ATR interface")
from the i2c tree and commits:
24868501a744 ("media: i2c: ds90ub9xx: Add err parameter to read/write funcs")
2ca499384e98 ("media: i2c: ds90ub960: Add RX port iteration support")
from the v4l-dvb tree.
I fixed it up (see below) and can carry the fix as necessary. This
is now fixed as far as linux-next is concerned, but any non trivial
conflicts should be mentioned to your upstream maintainer when your tree
is submitted for merging. You may also want to consider cooperating
with the maintainer of the conflicting tree to minimise any particularly
complex conflicts.
--
Cheers,
Stephen Rothwell
diff --cc drivers/media/i2c/ds90ub960.c
index 5a4d5de110bd,1877eb735cc7..000000000000
--- a/drivers/media/i2c/ds90ub960.c
+++ b/drivers/media/i2c/ds90ub960.c
@@@ -1056,11 -1271,10 +1274,12 @@@ static int ub960_atr_attach_addr(struc
struct ub960_rxport *rxport = priv->rxports[chan_id];
struct device *dev = &priv->client->dev;
unsigned int reg_idx;
+ int ret = 0;
- for (reg_idx = 0; reg_idx < ARRAY_SIZE(rxport->aliased_clients); reg_idx++) {
- if (!rxport->aliased_clients[reg_idx])
+ guard(mutex)(&rxport->aliased_addrs_lock);
+
+ for (reg_idx = 0; reg_idx < ARRAY_SIZE(rxport->aliased_addrs); reg_idx++) {
+ if (!rxport->aliased_addrs[reg_idx])
break;
}
@@@ -1069,15 -1283,18 +1288,18 @@@
return -EADDRNOTAVAIL;
}
- rxport->aliased_clients[reg_idx] = client;
+ rxport->aliased_addrs[reg_idx] = addr;
ub960_rxport_write(priv, chan_id, UB960_RR_SLAVE_ID(reg_idx),
- addr << 1);
- client->addr << 1, &ret);
++ addr << 1, &ret);
ub960_rxport_write(priv, chan_id, UB960_RR_SLAVE_ALIAS(reg_idx),
- alias << 1);
+ alias << 1, &ret);
+
+ if (ret)
+ return ret;
dev_dbg(dev, "rx%u: client 0x%02x assigned alias 0x%02x at slot %u\n",
- rxport->nport, client->addr, alias, reg_idx);
+ rxport->nport, addr, alias, reg_idx);
return 0;
}
@@@ -1089,11 -1306,10 +1311,12 @@@ static void ub960_atr_detach_addr(struc
struct ub960_rxport *rxport = priv->rxports[chan_id];
struct device *dev = &priv->client->dev;
unsigned int reg_idx;
+ int ret;
- for (reg_idx = 0; reg_idx < ARRAY_SIZE(rxport->aliased_clients); reg_idx++) {
- if (rxport->aliased_clients[reg_idx] == client)
+ guard(mutex)(&rxport->aliased_addrs_lock);
+
+ for (reg_idx = 0; reg_idx < ARRAY_SIZE(rxport->aliased_addrs); reg_idx++) {
+ if (rxport->aliased_addrs[reg_idx] == addr)
break;
}
@@@ -1103,12 -1319,18 +1326,18 @@@
return;
}
- rxport->aliased_clients[reg_idx] = NULL;
+ rxport->aliased_addrs[reg_idx] = 0;
- ub960_rxport_write(priv, chan_id, UB960_RR_SLAVE_ALIAS(reg_idx), 0);
+ ret = ub960_rxport_write(priv, chan_id, UB960_RR_SLAVE_ALIAS(reg_idx),
+ 0, NULL);
+ if (ret) {
+ dev_err(dev, "rx%u: unable to fully unmap client 0x%02x: %d\n",
+ rxport->nport, client->addr, ret);
+ return;
+ }
dev_dbg(dev, "rx%u: client 0x%02x released at slot %u\n", rxport->nport,
- client->addr, reg_idx);
+ addr, reg_idx);
}
static const struct i2c_atr_ops ub960_atr_ops = {
@@@ -3231,21 -4370,12 +4376,14 @@@ static void ub960_txport_free_ports(str
static void ub960_rxport_free_ports(struct ub960_data *priv)
{
- unsigned int nport;
-
- for (nport = 0; nport < priv->hw_data->num_rxports; nport++) {
- struct ub960_rxport *rxport = priv->rxports[nport];
-
- if (!rxport)
- continue;
-
- fwnode_handle_put(rxport->source.ep_fwnode);
- fwnode_handle_put(rxport->ser.fwnode);
+ for_each_active_rxport(priv, it) {
+ fwnode_handle_put(it.rxport->source.ep_fwnode);
+ fwnode_handle_put(it.rxport->ser.fwnode);
+ mutex_destroy(&rxport->aliased_addrs_lock);
+
- kfree(rxport);
- priv->rxports[nport] = NULL;
+ kfree(it.rxport);
+ priv->rxports[it.nport] = NULL;
}
}
Hi all,
On Mon, 28 Apr 2025 10:49:05 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>
> Today's linux-next merge of the v4l-dvb tree got a conflict in:
>
> drivers/media/i2c/ds90ub960.c
>
> between commits:
>
> 3ec29d51b546 ("media: i2c: ds90ub960: Protect alias_use_mask with a mutex")
> 818bd489f137 ("i2c: use client addresses directly in ATR interface")
>
> from the i2c tree and commits:
>
> 24868501a744 ("media: i2c: ds90ub9xx: Add err parameter to read/write funcs")
> 2ca499384e98 ("media: i2c: ds90ub960: Add RX port iteration support")
>
> from the v4l-dvb tree.
>
> I fixed it up (see below) and can carry the fix as necessary. This
> is now fixed as far as linux-next is concerned, but any non trivial
> conflicts should be mentioned to your upstream maintainer when your tree
> is submitted for merging. You may also want to consider cooperating
> with the maintainer of the conflicting tree to minimise any particularly
> complex conflicts.
The actual resolution is below ...
--
Cheers,
Stephen Rothwell
Hi all,
On Mon, 28 Apr 2025 11:22:00 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>
> On Mon, 28 Apr 2025 10:49:05 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> >
> > Today's linux-next merge of the v4l-dvb tree got a conflict in:
> >
> > drivers/media/i2c/ds90ub960.c
> >
> > between commits:
> >
> > 3ec29d51b546 ("media: i2c: ds90ub960: Protect alias_use_mask with a mutex")
> > 818bd489f137 ("i2c: use client addresses directly in ATR interface")
> >
> > from the i2c tree and commits:
> >
> > 24868501a744 ("media: i2c: ds90ub9xx: Add err parameter to read/write funcs")
> > 2ca499384e98 ("media: i2c: ds90ub960: Add RX port iteration support")
> >
> > from the v4l-dvb tree.
> >
> > I fixed it up (see below) and can carry the fix as necessary. This
> > is now fixed as far as linux-next is concerned, but any non trivial
> > conflicts should be mentioned to your upstream maintainer when your tree
> > is submitted for merging. You may also want to consider cooperating
> > with the maintainer of the conflicting tree to minimise any particularly
> > complex conflicts.
>
> The actual resolution is below ...
I hit the wrong key :-( Resolution below.
--
Cheers,
Stephen Rothwell
diff --cc drivers/media/i2c/ds90ub960.c
index 5a4d5de110bd,1877eb735cc7..000000000000
--- a/drivers/media/i2c/ds90ub960.c
+++ b/drivers/media/i2c/ds90ub960.c
@@@ -1056,11 -1271,10 +1274,12 @@@ static int ub960_atr_attach_addr(struc
struct ub960_rxport *rxport = priv->rxports[chan_id];
struct device *dev = &priv->client->dev;
unsigned int reg_idx;
+ int ret = 0;
- for (reg_idx = 0; reg_idx < ARRAY_SIZE(rxport->aliased_clients); reg_idx++) {
- if (!rxport->aliased_clients[reg_idx])
+ guard(mutex)(&rxport->aliased_addrs_lock);
+
+ for (reg_idx = 0; reg_idx < ARRAY_SIZE(rxport->aliased_addrs); reg_idx++) {
+ if (!rxport->aliased_addrs[reg_idx])
break;
}
@@@ -1069,15 -1283,18 +1288,18 @@@
return -EADDRNOTAVAIL;
}
- rxport->aliased_clients[reg_idx] = client;
+ rxport->aliased_addrs[reg_idx] = addr;
ub960_rxport_write(priv, chan_id, UB960_RR_SLAVE_ID(reg_idx),
- addr << 1);
- client->addr << 1, &ret);
++ addr << 1, &ret);
ub960_rxport_write(priv, chan_id, UB960_RR_SLAVE_ALIAS(reg_idx),
- alias << 1);
+ alias << 1, &ret);
+
+ if (ret)
+ return ret;
dev_dbg(dev, "rx%u: client 0x%02x assigned alias 0x%02x at slot %u\n",
- rxport->nport, client->addr, alias, reg_idx);
+ rxport->nport, addr, alias, reg_idx);
return 0;
}
@@@ -1089,11 -1306,10 +1311,12 @@@ static void ub960_atr_detach_addr(struc
struct ub960_rxport *rxport = priv->rxports[chan_id];
struct device *dev = &priv->client->dev;
unsigned int reg_idx;
+ int ret;
- for (reg_idx = 0; reg_idx < ARRAY_SIZE(rxport->aliased_clients); reg_idx++) {
- if (rxport->aliased_clients[reg_idx] == client)
+ guard(mutex)(&rxport->aliased_addrs_lock);
+
+ for (reg_idx = 0; reg_idx < ARRAY_SIZE(rxport->aliased_addrs); reg_idx++) {
+ if (rxport->aliased_addrs[reg_idx] == addr)
break;
}
@@@ -1103,12 -1319,18 +1326,18 @@@
return;
}
- rxport->aliased_clients[reg_idx] = NULL;
+ rxport->aliased_addrs[reg_idx] = 0;
- ub960_rxport_write(priv, chan_id, UB960_RR_SLAVE_ALIAS(reg_idx), 0);
+ ret = ub960_rxport_write(priv, chan_id, UB960_RR_SLAVE_ALIAS(reg_idx),
+ 0, NULL);
+ if (ret) {
+ dev_err(dev, "rx%u: unable to fully unmap client 0x%02x: %d\n",
- rxport->nport, client->addr, ret);
++ rxport->nport, addr, ret);
+ return;
+ }
dev_dbg(dev, "rx%u: client 0x%02x released at slot %u\n", rxport->nport,
- client->addr, reg_idx);
+ addr, reg_idx);
}
static const struct i2c_atr_ops ub960_atr_ops = {
@@@ -3231,21 -4370,12 +4376,14 @@@ static void ub960_txport_free_ports(str
static void ub960_rxport_free_ports(struct ub960_data *priv)
{
- unsigned int nport;
+ for_each_active_rxport(priv, it) {
+ fwnode_handle_put(it.rxport->source.ep_fwnode);
+ fwnode_handle_put(it.rxport->ser.fwnode);
- for (nport = 0; nport < priv->hw_data->num_rxports; nport++) {
- struct ub960_rxport *rxport = priv->rxports[nport];
++ mutex_destroy(&it.rxport->aliased_addrs_lock);
+
- if (!rxport)
- continue;
-
- fwnode_handle_put(rxport->source.ep_fwnode);
- fwnode_handle_put(rxport->ser.fwnode);
-
- mutex_destroy(&rxport->aliased_addrs_lock);
-
- kfree(rxport);
- priv->rxports[nport] = NULL;
+ kfree(it.rxport);
+ priv->rxports[it.nport] = NULL;
}
}
Hi all,
On Mon, 28 Apr 2025 11:30:52 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>
> On Mon, 28 Apr 2025 11:22:00 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> >
> > On Mon, 28 Apr 2025 10:49:05 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> > >
> > > Today's linux-next merge of the v4l-dvb tree got a conflict in:
> > >
> > > drivers/media/i2c/ds90ub960.c
> > >
> > > between commits:
> > >
> > > 3ec29d51b546 ("media: i2c: ds90ub960: Protect alias_use_mask with a mutex")
> > > 818bd489f137 ("i2c: use client addresses directly in ATR interface")
> > >
> > > from the i2c tree and commits:
> > >
> > > 24868501a744 ("media: i2c: ds90ub9xx: Add err parameter to read/write funcs")
> > > 2ca499384e98 ("media: i2c: ds90ub960: Add RX port iteration support")
> > >
> > > from the v4l-dvb tree.
> > >
> > > I fixed it up (see below) and can carry the fix as necessary. This
> > > is now fixed as far as linux-next is concerned, but any non trivial
> > > conflicts should be mentioned to your upstream maintainer when your tree
> > > is submitted for merging. You may also want to consider cooperating
> > > with the maintainer of the conflicting tree to minimise any particularly
> > > complex conflicts.
> >
> > The actual resolution is below ...
>
> I hit the wrong key :-( Resolution below.
> --
> Cheers,
> Stephen Rothwell
>
> diff --cc drivers/media/i2c/ds90ub960.c
> index 5a4d5de110bd,1877eb735cc7..000000000000
> --- a/drivers/media/i2c/ds90ub960.c
> +++ b/drivers/media/i2c/ds90ub960.c
> @@@ -1056,11 -1271,10 +1274,12 @@@ static int ub960_atr_attach_addr(struc
> struct ub960_rxport *rxport = priv->rxports[chan_id];
> struct device *dev = &priv->client->dev;
> unsigned int reg_idx;
> + int ret = 0;
>
> - for (reg_idx = 0; reg_idx < ARRAY_SIZE(rxport->aliased_clients); reg_idx++) {
> - if (!rxport->aliased_clients[reg_idx])
> + guard(mutex)(&rxport->aliased_addrs_lock);
> +
> + for (reg_idx = 0; reg_idx < ARRAY_SIZE(rxport->aliased_addrs); reg_idx++) {
> + if (!rxport->aliased_addrs[reg_idx])
> break;
> }
>
> @@@ -1069,15 -1283,18 +1288,18 @@@
> return -EADDRNOTAVAIL;
> }
>
> - rxport->aliased_clients[reg_idx] = client;
> + rxport->aliased_addrs[reg_idx] = addr;
>
> ub960_rxport_write(priv, chan_id, UB960_RR_SLAVE_ID(reg_idx),
> - addr << 1);
> - client->addr << 1, &ret);
> ++ addr << 1, &ret);
> ub960_rxport_write(priv, chan_id, UB960_RR_SLAVE_ALIAS(reg_idx),
> - alias << 1);
> + alias << 1, &ret);
> +
> + if (ret)
> + return ret;
>
> dev_dbg(dev, "rx%u: client 0x%02x assigned alias 0x%02x at slot %u\n",
> - rxport->nport, client->addr, alias, reg_idx);
> + rxport->nport, addr, alias, reg_idx);
>
> return 0;
> }
> @@@ -1089,11 -1306,10 +1311,12 @@@ static void ub960_atr_detach_addr(struc
> struct ub960_rxport *rxport = priv->rxports[chan_id];
> struct device *dev = &priv->client->dev;
> unsigned int reg_idx;
> + int ret;
>
> - for (reg_idx = 0; reg_idx < ARRAY_SIZE(rxport->aliased_clients); reg_idx++) {
> - if (rxport->aliased_clients[reg_idx] == client)
> + guard(mutex)(&rxport->aliased_addrs_lock);
> +
> + for (reg_idx = 0; reg_idx < ARRAY_SIZE(rxport->aliased_addrs); reg_idx++) {
> + if (rxport->aliased_addrs[reg_idx] == addr)
> break;
> }
>
> @@@ -1103,12 -1319,18 +1326,18 @@@
> return;
> }
>
> - rxport->aliased_clients[reg_idx] = NULL;
> + rxport->aliased_addrs[reg_idx] = 0;
>
> - ub960_rxport_write(priv, chan_id, UB960_RR_SLAVE_ALIAS(reg_idx), 0);
> + ret = ub960_rxport_write(priv, chan_id, UB960_RR_SLAVE_ALIAS(reg_idx),
> + 0, NULL);
> + if (ret) {
> + dev_err(dev, "rx%u: unable to fully unmap client 0x%02x: %d\n",
> - rxport->nport, client->addr, ret);
> ++ rxport->nport, addr, ret);
> + return;
> + }
>
> dev_dbg(dev, "rx%u: client 0x%02x released at slot %u\n", rxport->nport,
> - client->addr, reg_idx);
> + addr, reg_idx);
> }
>
> static const struct i2c_atr_ops ub960_atr_ops = {
> @@@ -3231,21 -4370,12 +4376,14 @@@ static void ub960_txport_free_ports(str
>
> static void ub960_rxport_free_ports(struct ub960_data *priv)
> {
> - unsigned int nport;
> + for_each_active_rxport(priv, it) {
> + fwnode_handle_put(it.rxport->source.ep_fwnode);
> + fwnode_handle_put(it.rxport->ser.fwnode);
>
> - for (nport = 0; nport < priv->hw_data->num_rxports; nport++) {
> - struct ub960_rxport *rxport = priv->rxports[nport];
> ++ mutex_destroy(&it.rxport->aliased_addrs_lock);
> +
> - if (!rxport)
> - continue;
> -
> - fwnode_handle_put(rxport->source.ep_fwnode);
> - fwnode_handle_put(rxport->ser.fwnode);
> -
> - mutex_destroy(&rxport->aliased_addrs_lock);
> -
> - kfree(rxport);
> - priv->rxports[nport] = NULL;
> + kfree(it.rxport);
> + priv->rxports[it.nport] = NULL;
> }
> }
This is now a conflict between the i2c tree and Linus' tree.
--
Cheers,
Stephen Rothwell
Hi Stephen,
On Thursday, 29 May 2025 04:49:29 CEST Stephen Rothwell wrote:
> Hi all,
>
> On Mon, 28 Apr 2025 11:30:52 +1000 Stephen Rothwell <sfr@canb.auug.org.au>
wrote:
> > On Mon, 28 Apr 2025 11:22:00 +1000 Stephen Rothwell <sfr@canb.auug.org.au>
wrote:
> > > On Mon, 28 Apr 2025 10:49:05 +1000 Stephen Rothwell
<sfr@canb.auug.org.au> wrote:
> > > > Today's linux-next merge of the v4l-dvb tree got a conflict in:
> > > > drivers/media/i2c/ds90ub960.c
> > > >
> > > > between commits:
> > > > 3ec29d51b546 ("media: i2c: ds90ub960: Protect alias_use_mask with a
> > > > mutex")
> > > > 818bd489f137 ("i2c: use client addresses directly in ATR interface")
> > > >
> > > > from the i2c tree and commits:
> > > > 24868501a744 ("media: i2c: ds90ub9xx: Add err parameter to
> > > > read/write funcs") 2ca499384e98 ("media: i2c: ds90ub960: Add RX
> > > > port iteration support")> > >
> > > > from the v4l-dvb tree.
> > > >
> > > > I fixed it up (see below) and can carry the fix as necessary. This
> > > > is now fixed as far as linux-next is concerned, but any non trivial
> > > > conflicts should be mentioned to your upstream maintainer when your
> > > > tree
> > > > is submitted for merging. You may also want to consider cooperating
> > > > with the maintainer of the conflicting tree to minimise any
> > > > particularly
> > > > complex conflicts.
> > >
> > > The actual resolution is below ...
> >
> > I hit the wrong key :-( Resolution below.
>
> This is now a conflict between the i2c tree and Linus' tree.
Below is the resolution I came up with.
Thanks,
--
Romain Gantois, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
diff --cc drivers/media/i2c/ds90ub960.c
index ed9ace1a54766,94b20ba6cb86f..54c2546551451
--- a/drivers/media/i2c/ds90ub960.c
+++ b/drivers/media/i2c/ds90ub960.c
@@@ -1271,10 -1056,11 +1274,12 @@@ static int ub960_atr_attach_addr(struc
struct ub960_rxport *rxport = priv->rxports[chan_id];
struct device *dev = &priv->client->dev;
unsigned int reg_idx;
+ int ret = 0;
- for (reg_idx = 0; reg_idx < ARRAY_SIZE(rxport->aliased_clients);
reg_idx++) {
- if (!rxport->aliased_clients[reg_idx])
+ guard(mutex)(&rxport->aliased_addrs_lock);
+
+ for (reg_idx = 0; reg_idx < ARRAY_SIZE(rxport->aliased_addrs);
reg_idx++) {
+ if (!rxport->aliased_addrs[reg_idx])
break;
}
@@@ -1283,18 -1069,15 +1288,18 @@@
return -EADDRNOTAVAIL;
}
- rxport->aliased_clients[reg_idx] = client;
+ rxport->aliased_addrs[reg_idx] = addr;
ub960_rxport_write(priv, chan_id, UB960_RR_SLAVE_ID(reg_idx),
- client->addr << 1, &ret);
- addr << 1);
++ addr << 1, &ret);
ub960_rxport_write(priv, chan_id, UB960_RR_SLAVE_ALIAS(reg_idx),
- alias << 1);
+ alias << 1, &ret);
+
+ if (ret)
+ return ret;
dev_dbg(dev, "rx%u: client 0x%02x assigned alias 0x%02x at slot %u\n",
- rxport->nport, client->addr, alias, reg_idx);
+ rxport->nport, addr, alias, reg_idx);
return 0;
}
@@@ -1306,10 -1089,11 +1311,12 @@@ static void ub960_atr_detach_addr(struc
struct ub960_rxport *rxport = priv->rxports[chan_id];
struct device *dev = &priv->client->dev;
unsigned int reg_idx;
+ int ret;
- for (reg_idx = 0; reg_idx < ARRAY_SIZE(rxport->aliased_clients);
reg_idx++) {
- if (rxport->aliased_clients[reg_idx] == client)
+ guard(mutex)(&rxport->aliased_addrs_lock);
+
+ for (reg_idx = 0; reg_idx < ARRAY_SIZE(rxport->aliased_addrs);
reg_idx++) {
+ if (rxport->aliased_addrs[reg_idx] == addr)
break;
}
@@@ -1319,18 -1103,12 +1326,18 @@@
return;
}
- rxport->aliased_clients[reg_idx] = NULL;
+ rxport->aliased_addrs[reg_idx] = 0;
- ub960_rxport_write(priv, chan_id, UB960_RR_SLAVE_ALIAS(reg_idx), 0);
+ ret = ub960_rxport_write(priv, chan_id, UB960_RR_SLAVE_ALIAS(reg_idx),
+ 0, NULL);
+ if (ret) {
+ dev_err(dev, "rx%u: unable to fully unmap client 0x%02x:
%d\n",
+ rxport->nport, client->addr, ret);
+ return;
+ }
dev_dbg(dev, "rx%u: client 0x%02x released at slot %u\n", rxport-
>nport,
- client->addr, reg_idx);
+ addr, reg_idx);
}
static const struct i2c_atr_ops ub960_atr_ops = {
@@@ -4370,12 -3231,21 +4376,14 @@@ static void ub960_txport_free_ports(str
static void ub960_rxport_free_ports(struct ub960_data *priv)
{
- unsigned int nport;
-
- for (nport = 0; nport < priv->hw_data->num_rxports; nport++) {
- struct ub960_rxport *rxport = priv->rxports[nport];
-
- if (!rxport)
- continue;
-
- fwnode_handle_put(rxport->source.ep_fwnode);
- fwnode_handle_put(rxport->ser.fwnode);
+ for_each_active_rxport(priv, it) {
+ fwnode_handle_put(it.rxport->source.ep_fwnode);
+ fwnode_handle_put(it.rxport->ser.fwnode);
+ mutex_destroy(&rxport->aliased_addrs_lock);
+
- kfree(rxport);
- priv->rxports[nport] = NULL;
+ kfree(it.rxport);
+ priv->rxports[it.nport] = NULL;
}
}
> Below is the resolution I came up with. Linus solved it differently [1]. I think he is right, but those interested please double check. [1] https://lore.kernel.org/all/CAHk-=wiKW=BPcDvBAsVDemdWBR0uh09A_WMOCoceqj3w3doGJg@mail.gmail.com/
Hi Wolfram,
On Saturday, 31 May 2025 14:05:03 CEST Wolfram Sang wrote:
> > Below is the resolution I came up with.
>
> Linus solved it differently [1]. I think he is right, but those
> interested please double check.
>
> [1]
> https://lore.kernel.org/all/CAHk-=wiKW=BPcDvBAsVDemdWBR0uh09A_WMOCoceqj3w3d
> oGJg@mail.gmail.com/
Indeed Linus' resolution is correct, the mutex destroy line should be:
mutex_destroy(&it.rxport->aliased_addrs_lock);
Thanks,
--
Romain Gantois, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
> Below is the resolution I came up with.
Thank you for this...
> - for (reg_idx = 0; reg_idx < ARRAY_SIZE(rxport->aliased_clients);
> reg_idx++) {
... but it was mangled, sadly. Can you resend it properly, please?
Hi Wolfram,
On Friday, 30 May 2025 13:24:46 CEST Wolfram Sang wrote:
> > Below is the resolution I came up with.
>
> Thank you for this...
>
> > - for (reg_idx = 0; reg_idx < ARRAY_SIZE(rxport->aliased_clients);
> > reg_idx++) {
>
> ... but it was mangled, sadly. Can you resend it properly, please?
Sorry about that, here's a hopefully properly formatted version. Besides the
issue that Linus found with the mutex_destroy(), I'd also missed a
"s/client->addr/addr" in the previous version.
By the way, this is just the output of "git show" after resolving the merge,
I'm not used to sending these "resolutions diffs" so please let me know
if you need it in some other form.
Thanks,
---
diff --cc drivers/media/i2c/ds90ub960.c
index ed9ace1a54766,94b20ba6cb86f..082fc62b0f5b9
--- a/drivers/media/i2c/ds90ub960.c
+++ b/drivers/media/i2c/ds90ub960.c
@@@ -1271,10 -1056,11 +1274,12 @@@ static int ub960_atr_attach_addr(struc
struct ub960_rxport *rxport = priv->rxports[chan_id];
struct device *dev = &priv->client->dev;
unsigned int reg_idx;
+ int ret = 0;
- for (reg_idx = 0; reg_idx < ARRAY_SIZE(rxport->aliased_clients); reg_idx++) {
- if (!rxport->aliased_clients[reg_idx])
+ guard(mutex)(&rxport->aliased_addrs_lock);
+
+ for (reg_idx = 0; reg_idx < ARRAY_SIZE(rxport->aliased_addrs); reg_idx++) {
+ if (!rxport->aliased_addrs[reg_idx])
break;
}
@@@ -1283,18 -1069,15 +1288,18 @@@
return -EADDRNOTAVAIL;
}
- rxport->aliased_clients[reg_idx] = client;
+ rxport->aliased_addrs[reg_idx] = addr;
ub960_rxport_write(priv, chan_id, UB960_RR_SLAVE_ID(reg_idx),
- client->addr << 1, &ret);
- addr << 1);
++ addr << 1, &ret);
ub960_rxport_write(priv, chan_id, UB960_RR_SLAVE_ALIAS(reg_idx),
- alias << 1);
+ alias << 1, &ret);
+
+ if (ret)
+ return ret;
dev_dbg(dev, "rx%u: client 0x%02x assigned alias 0x%02x at slot %u\n",
- rxport->nport, client->addr, alias, reg_idx);
+ rxport->nport, addr, alias, reg_idx);
return 0;
}
@@@ -1306,10 -1089,11 +1311,12 @@@ static void ub960_atr_detach_addr(struc
struct ub960_rxport *rxport = priv->rxports[chan_id];
struct device *dev = &priv->client->dev;
unsigned int reg_idx;
+ int ret;
- for (reg_idx = 0; reg_idx < ARRAY_SIZE(rxport->aliased_clients); reg_idx++) {
- if (rxport->aliased_clients[reg_idx] == client)
+ guard(mutex)(&rxport->aliased_addrs_lock);
+
+ for (reg_idx = 0; reg_idx < ARRAY_SIZE(rxport->aliased_addrs); reg_idx++) {
+ if (rxport->aliased_addrs[reg_idx] == addr)
break;
}
@@@ -1319,18 -1103,12 +1326,18 @@@
return;
}
- rxport->aliased_clients[reg_idx] = NULL;
+ rxport->aliased_addrs[reg_idx] = 0;
- ub960_rxport_write(priv, chan_id, UB960_RR_SLAVE_ALIAS(reg_idx), 0);
+ ret = ub960_rxport_write(priv, chan_id, UB960_RR_SLAVE_ALIAS(reg_idx),
+ 0, NULL);
+ if (ret) {
+ dev_err(dev, "rx%u: unable to fully unmap client 0x%02x: %d\n",
- rxport->nport, client->addr, ret);
++ rxport->nport, addr, ret);
+ return;
+ }
dev_dbg(dev, "rx%u: client 0x%02x released at slot %u\n", rxport->nport,
- client->addr, reg_idx);
+ addr, reg_idx);
}
static const struct i2c_atr_ops ub960_atr_ops = {
@@@ -4370,12 -3231,21 +4376,14 @@@ static void ub960_txport_free_ports(str
static void ub960_rxport_free_ports(struct ub960_data *priv)
{
- unsigned int nport;
-
- for (nport = 0; nport < priv->hw_data->num_rxports; nport++) {
- struct ub960_rxport *rxport = priv->rxports[nport];
-
- if (!rxport)
- continue;
-
- fwnode_handle_put(rxport->source.ep_fwnode);
- fwnode_handle_put(rxport->ser.fwnode);
+ for_each_active_rxport(priv, it) {
+ fwnode_handle_put(it.rxport->source.ep_fwnode);
+ fwnode_handle_put(it.rxport->ser.fwnode);
- mutex_destroy(&rxport->aliased_addrs_lock);
++ mutex_destroy(&it.rxport->aliased_addrs_lock);
+
- kfree(rxport);
- priv->rxports[nport] = NULL;
+ kfree(it.rxport);
+ priv->rxports[it.nport] = NULL;
}
}
Hi,
On 28/04/2025 04:30, Stephen Rothwell wrote:
> Hi all,
>
> On Mon, 28 Apr 2025 11:22:00 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>>
>> On Mon, 28 Apr 2025 10:49:05 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>>>
>>> Today's linux-next merge of the v4l-dvb tree got a conflict in:
>>>
>>> drivers/media/i2c/ds90ub960.c
>>>
>>> between commits:
>>>
>>> 3ec29d51b546 ("media: i2c: ds90ub960: Protect alias_use_mask with a mutex")
>>> 818bd489f137 ("i2c: use client addresses directly in ATR interface")
>>>
>>> from the i2c tree and commits:
>>>
>>> 24868501a744 ("media: i2c: ds90ub9xx: Add err parameter to read/write funcs")
>>> 2ca499384e98 ("media: i2c: ds90ub960: Add RX port iteration support")
>>>
>>> from the v4l-dvb tree.
>>>
>>> I fixed it up (see below) and can carry the fix as necessary. This
>>> is now fixed as far as linux-next is concerned, but any non trivial
>>> conflicts should be mentioned to your upstream maintainer when your tree
>>> is submitted for merging. You may also want to consider cooperating
>>> with the maintainer of the conflicting tree to minimise any particularly
>>> complex conflicts.
>>
>> The actual resolution is below ...
>
> I hit the wrong key :-( Resolution below.
I came up with the same resolution, so looks correct to me.
Tomi
© 2016 - 2026 Red Hat, Inc.