Re: [PATCH] hw/net/imx_fec: return 0xffff when accessing non-existing PHY

[Guenter Roeck]

On 5/2/21 9:09 AM, Bin Meng wrote:
> On Mon, May 3, 2021 at 12:03 AM Guenter Roeck <linux@roeck-us.net> wrote:
>>
>> If a PHY does not exist, attempts to read from it should return 0xffff.
>> Otherwise the Linux kernel will believe that a PHY is there and select
>> the non-existing PHY. This in turn will result in network errors later
>> on since the real PHY is not selected or configured.
>>
>> Since reading from or writing to a non-existing PHY is not an emulation
>> error, replace guest error messages with traces.
>>
>> Fixes: 461c51ad4275 ("Add a phy-num property to the i.MX FEC emulator")
>> Cc: Jean-Christophe Dubois <jcd@tribudubois.net>
>> Signed-off-by: Guenter Roeck <linux@roeck-us.net>
>> ---
>>  hw/net/imx_fec.c    | 8 +++-----
>>  hw/net/trace-events | 2 ++
>>  2 files changed, 5 insertions(+), 5 deletions(-)
>>
>> diff --git a/hw/net/imx_fec.c b/hw/net/imx_fec.c
>> index f03450c028..9c7035bc94 100644
>> --- a/hw/net/imx_fec.c
>> +++ b/hw/net/imx_fec.c
>> @@ -283,9 +283,8 @@ static uint32_t imx_phy_read(IMXFECState *s, int reg)
>>      uint32_t phy = reg / 32;
>>
>>      if (phy != s->phy_num) {
>> -        qemu_log_mask(LOG_GUEST_ERROR, "[%s.phy]%s: Bad phy num %u\n",
>> -                      TYPE_IMX_FEC, __func__, phy);
>> -        return 0;
>> +        trace_imx_phy_read_num(phy, s->phy_num);
>> +        return 0xffff;
>>      }
>>
>>      reg %= 32;
>> @@ -345,8 +344,7 @@ static void imx_phy_write(IMXFECState *s, int reg, 
>> uint32_t val)
>>      uint32_t phy = reg / 32;
>>
>>      if (phy != s->phy_num) {
>> -        qemu_log_mask(LOG_GUEST_ERROR, "[%s.phy]%s: Bad phy num %u\n",
>> -                      TYPE_IMX_FEC, __func__, phy);
>> +        trace_imx_phy_write_num(phy, s->phy_num);
>>          return;
>>      }
>>
>> diff --git a/hw/net/trace-events b/hw/net/trace-events
>> index baf25ffa7e..ee77238d9e 100644
>> --- a/hw/net/trace-events
>> +++ b/hw/net/trace-events
>> @@ -413,8 +413,10 @@ i82596_set_multicast(uint16_t count) "Added %d 
>> multicast entries"
>>  i82596_channel_attention(void *s) "%p: Received CHANNEL ATTENTION"
>>
>>  # imx_fec.c
>> +imx_phy_read_num(int phy, int configured) "read request from unconfigured 
>> phy %d (configured %d)"
> 
> nits: could we put this below imx_phy_read(), like you put
> imx_phy_write_num after imx_phy_write?
> 

Sure, makes sense. I'll wait a bit for other feedback and then resend.

>>  imx_phy_read(uint32_t val, int phy, int reg) "0x%04"PRIx32" <= 
>> phy[%d].reg[%d]"
>>  imx_phy_write(uint32_t val, int phy, int reg) "0x%04"PRIx32" => 
>> phy[%d].reg[%d]"
>> +imx_phy_write_num(int phy, int configured) "write request to unconfigured 
>> phy %d (configured %d)"
>>  imx_phy_update_link(const char *s) "%s"
>>  imx_phy_reset(void) ""
>>  imx_fec_read_bd(uint64_t addr, int flags, int len, int data) "tx_bd 
>> 0x%"PRIx64" flags 0x%04x len %d data 0x%08x"
>> --
> 
> Otherwise,
> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
> Tested-by: Bin Meng <bmeng.cn@gmail.com>
> 

Thanks!

Guenter