From: Corey Minyard <minyard@acm.org>
To: Alistair Francis <alistair.francis@xilinx.com>
Cc: "qemu-devel@nongnu.org Developers" <qemu-devel@nongnu.org>,
cminyard@mvista.com, Peter Maydell <peter.maydell@linaro.org>,
Peter Crosthwaite <crosthwaite.peter@gmail.com>,
Kwon <hyun.kwon@xilinx.com>,
KONRAD Frederic <fred.konrad@greensocs.com>
Subject: Re: [Qemu-devel] [PATCH] i2c: Fix SMBus read transactions to avoid double events
Date: Tue, 28 Jun 2016 11:21:52 -0500 [thread overview]
Message-ID: <5772A420.7030203@acm.org> (raw)
In-Reply-To: <CAKmqyKOv1vWU_LGQnw4jy9co__UnVCkUzg3onP78VCodwtZHnw@mail.gmail.com>
On 06/27/2016 07:43 PM, Alistair Francis wrote:
> On Mon, Jun 27, 2016 at 3:04 PM, <minyard@acm.org> wrote:
>> From: Corey Minyard <cminyard@mvista.com>
>>
>> Change 2293c27faddf (i2c: implement broadcast write) added broadcast
>> capability to the I2C bus, but it broke SMBus read transactions.
>> An SMBus read transaction does two i2c_start_transaction() calls
>> without an intervening i2c_end_transfer() call. This will
>> result in i2c_start_transfer() adding the same device to the
>> current_devs list twice, and then the ->event() for the same
>> device gets called twice in the second call to i2c_start_transfer(),
>> resulting in the smbus code getting confused.
>>
>> This fix adds a third state to the i2c_start_transfer() recv
>> parameter, a read continued that will not scan for devices
>> and add them to current_devs. It also adds #defines for all
>> the values for the recv parameter.
>>
>> This also deletes the empty check from the top of i2c_end_transfer().
>> It's unnecessary, and it prevents the broadcast variable from being
>> set to false at the end of the transaction if no devices were on
>> the bus.
>>
>> Cc: KONRAD Frederic <fred.konrad@greensocs.com>
>> Cc: Alistair Francis <alistair.francis@xilinx.com>
>> Cc: Peter Crosthwaite <crosthwaite.peter@gmail.com>
>> Cc: Kwon <hyun.kwon@xilinx.com>
>> Cc: Peter Maydell <peter.maydell@linaro.org>
>> Signed-off-by: Corey Minyard <cminyard@mvista.com>
>> ---
>> hw/i2c/core.c | 24 +++++++++++-------------
>> hw/i2c/smbus.c | 22 +++++++++++-----------
>> include/hw/i2c/i2c.h | 9 +++++++++
>> 3 files changed, 31 insertions(+), 24 deletions(-)
>>
>> I considered a couple of ways to do this. I thought about adding a
>> separate function to do a "intermediate end" of the transaction, but
>> that seemed like too much work. I also thought about adding a
>> bool saing whether you are currently in a transaction and not rescan
>> the bus if you are. However, that would require that the bool be in
>> the vmstate, and that would be complicated.
>>
>> On that note, the current_devs list is not in the vmstate. That means
>> that a migrate done in the middle of an I2C transaction will cause the
>> I2C transaction to fail, right? Maybe this whole broadcast thing is
>> a bad idea, or needs a different implementation?
Looking at it a little closer, migration does appear to be handled
correctly. So I'll stick with this patch.
>> diff --git a/hw/i2c/core.c b/hw/i2c/core.c
>> index abb3efb..53069dd 100644
>> --- a/hw/i2c/core.c
>> +++ b/hw/i2c/core.c
>> @@ -101,15 +101,17 @@ int i2c_start_transfer(I2CBus *bus, uint8_t address, int recv)
>> bus->broadcast = true;
>> }
>>
>> - QTAILQ_FOREACH(kid, &bus->qbus.children, sibling) {
>> - DeviceState *qdev = kid->child;
>> - I2CSlave *candidate = I2C_SLAVE(qdev);
>> - if ((candidate->address == address) || (bus->broadcast)) {
>> - node = g_malloc(sizeof(struct I2CNode));
>> - node->elt = candidate;
>> - QLIST_INSERT_HEAD(&bus->current_devs, node, next);
>> - if (!bus->broadcast) {
>> - break;
>> + if (recv != I2C_START_CONTINUED_READ_TRANSFER) {
>> + QTAILQ_FOREACH(kid, &bus->qbus.children, sibling) {
>> + DeviceState *qdev = kid->child;
>> + I2CSlave *candidate = I2C_SLAVE(qdev);
>> + if ((candidate->address == address) || (bus->broadcast)) {
>> + node = g_malloc(sizeof(struct I2CNode));
>> + node->elt = candidate;
>> + QLIST_INSERT_HEAD(&bus->current_devs, node, next);
>> + if (!bus->broadcast) {
>> + break;
>> + }
>> }
>> }
>> }
>> @@ -134,10 +136,6 @@ void i2c_end_transfer(I2CBus *bus)
>> I2CSlaveClass *sc;
>> I2CNode *node, *next;
>>
>> - if (QLIST_EMPTY(&bus->current_devs)) {
>> - return;
>> - }
>> -
>> QLIST_FOREACH_SAFE(node, &bus->current_devs, next, next) {
>> sc = I2C_SLAVE_GET_CLASS(node->elt);
>> if (sc->event) {
>> diff --git a/hw/i2c/smbus.c b/hw/i2c/smbus.c
>> index 3979b3d..f63799d 100644
>> --- a/hw/i2c/smbus.c
>> +++ b/hw/i2c/smbus.c
>> @@ -222,7 +222,7 @@ int smbus_receive_byte(I2CBus *bus, uint8_t addr)
>> {
>> uint8_t data;
>>
>> - if (i2c_start_transfer(bus, addr, 1)) {
>> + if (i2c_start_transfer(bus, addr, I2C_START_READ_TRANSFER)) {
>> return -1;
>> }
>> data = i2c_recv(bus);
>> @@ -233,7 +233,7 @@ int smbus_receive_byte(I2CBus *bus, uint8_t addr)
>>
>> int smbus_send_byte(I2CBus *bus, uint8_t addr, uint8_t data)
>> {
>> - if (i2c_start_transfer(bus, addr, 0)) {
>> + if (i2c_start_transfer(bus, addr, I2C_START_WRITE_TRANSFER)) {
>> return -1;
>> }
>> i2c_send(bus, data);
>> @@ -244,11 +244,11 @@ int smbus_send_byte(I2CBus *bus, uint8_t addr, uint8_t data)
>> int smbus_read_byte(I2CBus *bus, uint8_t addr, uint8_t command)
>> {
>> uint8_t data;
>> - if (i2c_start_transfer(bus, addr, 0)) {
>> + if (i2c_start_transfer(bus, addr, I2C_START_WRITE_TRANSFER)) {
>> return -1;
>> }
>> i2c_send(bus, command);
>> - i2c_start_transfer(bus, addr, 1);
>> + i2c_start_transfer(bus, addr, I2C_START_CONTINUED_READ_TRANSFER);
>> data = i2c_recv(bus);
>> i2c_nack(bus);
>> i2c_end_transfer(bus);
>> @@ -257,7 +257,7 @@ int smbus_read_byte(I2CBus *bus, uint8_t addr, uint8_t command)
>>
>> int smbus_write_byte(I2CBus *bus, uint8_t addr, uint8_t command, uint8_t data)
>> {
>> - if (i2c_start_transfer(bus, addr, 0)) {
>> + if (i2c_start_transfer(bus, addr, I2C_START_WRITE_TRANSFER)) {
>> return -1;
>> }
>> i2c_send(bus, command);
>> @@ -269,11 +269,11 @@ int smbus_write_byte(I2CBus *bus, uint8_t addr, uint8_t command, uint8_t data)
>> int smbus_read_word(I2CBus *bus, uint8_t addr, uint8_t command)
>> {
>> uint16_t data;
>> - if (i2c_start_transfer(bus, addr, 0)) {
>> + if (i2c_start_transfer(bus, addr, I2C_START_WRITE_TRANSFER)) {
>> return -1;
>> }
>> i2c_send(bus, command);
>> - i2c_start_transfer(bus, addr, 1);
>> + i2c_start_transfer(bus, addr, I2C_START_CONTINUED_READ_TRANSFER);
>> data = i2c_recv(bus);
>> data |= i2c_recv(bus) << 8;
>> i2c_nack(bus);
>> @@ -283,7 +283,7 @@ int smbus_read_word(I2CBus *bus, uint8_t addr, uint8_t command)
>>
>> int smbus_write_word(I2CBus *bus, uint8_t addr, uint8_t command, uint16_t data)
>> {
>> - if (i2c_start_transfer(bus, addr, 0)) {
>> + if (i2c_start_transfer(bus, addr, I2C_START_WRITE_TRANSFER)) {
>> return -1;
>> }
>> i2c_send(bus, command);
>> @@ -298,11 +298,11 @@ int smbus_read_block(I2CBus *bus, uint8_t addr, uint8_t command, uint8_t *data)
>> int len;
>> int i;
>>
>> - if (i2c_start_transfer(bus, addr, 0)) {
>> + if (i2c_start_transfer(bus, addr, I2C_START_WRITE_TRANSFER)) {
>> return -1;
>> }
>> i2c_send(bus, command);
>> - i2c_start_transfer(bus, addr, 1);
>> + i2c_start_transfer(bus, addr, I2C_START_CONTINUED_READ_TRANSFER);
>> len = i2c_recv(bus);
>> if (len > 32) {
>> len = 0;
>> @@ -323,7 +323,7 @@ int smbus_write_block(I2CBus *bus, uint8_t addr, uint8_t command, uint8_t *data,
>> if (len > 32)
>> len = 32;
>>
>> - if (i2c_start_transfer(bus, addr, 0)) {
>> + if (i2c_start_transfer(bus, addr, I2C_START_WRITE_TRANSFER)) {
>> return -1;
>> }
>> i2c_send(bus, command);
>> diff --git a/include/hw/i2c/i2c.h b/include/hw/i2c/i2c.h
>> index c4085aa..16c910e 100644
>> --- a/include/hw/i2c/i2c.h
>> +++ b/include/hw/i2c/i2c.h
>> @@ -50,6 +50,15 @@ struct I2CSlave
>> uint8_t address;
>> };
>>
>> +/* For the recv value in i2c_start_transfer. The first two values
>> + correspond to false and true for the recv value. The third is a
>> + special value that is used to tell i2c_start_transfer that this is
>> + a continuation of the previous transfer, so don't rescan the bus
>> + for devices to send to, continue with the current set of devices. */
> This comment is a little confusing, I don't think you need to explain
> what the read/write values correspond to but you clear up the
> explanation about the continued read value. Something like this I like
> is clearer:
>
> The continued read is a special value that is used to tell the
> i2c_start_transfer() function that this is a continuation of the
> previous transfer so we don't rescan the bus for devices to send to
> and instead just continue with the current set of devices.
I did think about doing it this way, but it seemed clearer to me the
other way. But I think you are right.
-corey
> Otherwise:
> Reviewed-by: Alistair Francis <alistair.francis@xilinx.com>
>
> Thanks,
>
> Alistair
>
>
>> +#define I2C_START_WRITE_TRANSFER 0
>> +#define I2C_START_READ_TRANSFER 1
>> +#define I2C_START_CONTINUED_READ_TRANSFER 2
>> +
>> I2CBus *i2c_init_bus(DeviceState *parent, const char *name);
>> void i2c_set_slave_address(I2CSlave *dev, uint8_t address);
>> int i2c_bus_busy(I2CBus *bus);
>> --
>> 2.7.4
>>
>>
next prev parent reply other threads:[~2016-06-28 16:22 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-06-27 22:04 [Qemu-devel] [PATCH] i2c: Fix SMBus read transactions to avoid double events minyard
2016-06-28 0:43 ` Alistair Francis
2016-06-28 16:21 ` Corey Minyard [this message]
2016-06-28 17:45 ` Corey Minyard
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=5772A420.7030203@acm.org \
--to=minyard@acm.org \
--cc=alistair.francis@xilinx.com \
--cc=cminyard@mvista.com \
--cc=crosthwaite.peter@gmail.com \
--cc=fred.konrad@greensocs.com \
--cc=hyun.kwon@xilinx.com \
--cc=peter.maydell@linaro.org \
--cc=qemu-devel@nongnu.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).