linux-kernel - Re: [PATCH] serial: parisc: GSC: fix build when PCI

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Mon, 14 Feb 2022 13:24:54 +0100
From:   Helge Deller <deller@....de>
To:     Randy Dunlap <rdunlap@...radead.org>, linux-kernel@...r.kernel.org
Cc:     kernel test robot <lkp@...el.com>,
        "James E.J. Bottomley" <James.Bottomley@...senPartnership.com>,
        linux-parisc@...r.kernel.org,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        linux-serial@...r.kernel.org, Jiri Slaby <jirislaby@...nel.org>,
        Johan Hovold <johan@...nel.org>
Subject: Re: [PATCH] serial: parisc: GSC: fix build when PCI_LBA is not set

On 2/14/22 13:05, Helge Deller wrote:
> On 2/14/22 01:15, Randy Dunlap wrote:
>> Hi,
>>
>> On 2/13/22 14:15, Helge Deller wrote:
>>> On 2/13/22 22:07, Randy Dunlap wrote:
>>>>
>>>>
>>>> On 2/13/22 12:35, Helge Deller wrote:
>>>>> Hi Randy,
>>>>>
>>>>> On 2/13/22 20:39, Randy Dunlap wrote:
>>>>>> There is a build error when using a kernel .config file from
>>>>>> 'kernel test robot' for a different build problem:
>>>>>>
>>>>>> hppa64-linux-ld: drivers/tty/serial/8250/8250_gsc.o: in function `.LC3':
>>>>>> (.data.rel.ro+0x18): undefined reference to `iosapic_serial_irq'
>>>>>>
>>>>>> when:
>>>>>>   CONFIG_GSC=y
>>>>>>   CONFIG_SERIO_GSCPS2=y
>>>>>>   CONFIG_SERIAL_8250_GSC=y
>>>>>>   CONFIG_PCI is not set
>>>>>>     and hence PCI_LBA is not set.
>>>>>>   IOSAPIC depends on PCI_LBA, so IOSAPIC is not set/enabled.
>>>>>>
>>>>>> Making SERIAL_8250_GSC depend on PCI_LBA prevents the build error.
>>>>>
>>>>> It maybe makes the build error go away, but ...
>>>>>
>>>>>> Signed-off-by: Randy Dunlap <rdunlap@...radead.org>
>>>>>> Reported-by: kernel test robot <lkp@...el.com>
>>>>>> Cc: "James E.J. Bottomley" <James.Bottomley@...senPartnership.com>
>>>>>> Cc: Helge Deller <deller@....de>
>>>>>> Cc: linux-parisc@...r.kernel.org
>>>>>> Cc: Greg Kroah-Hartman <gregkh@...uxfoundation.org>
>>>>>> Cc: linux-serial@...r.kernel.org
>>>>>> Cc: Jiri Slaby <jirislaby@...nel.org>
>>>>>> Cc: Johan Hovold <johan@...nel.org>
>>>>>> ---
>>>>>>  drivers/tty/serial/8250/Kconfig |    2 +-
>>>>>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>>>>>
>>>>>> --- linux-next-20220211.orig/drivers/tty/serial/8250/Kconfig
>>>>>> +++ linux-next-20220211/drivers/tty/serial/8250/Kconfig
>>>>>> @@ -118,7 +118,7 @@ config SERIAL_8250_CONSOLE
>>>>>>
>>>>>>  config SERIAL_8250_GSC
>>>>>>  	tristate
>>>>>> -	depends on SERIAL_8250 && GSC
>>>>>> +	depends on SERIAL_8250 && GSC && PCI_LBA
>>>>>>  	default SERIAL_8250
>>>>>
>>>>> The serial device is on the GSC bus, so if you make it
>>>>> dependend on the PCI bus it will not be useable on machines
>>>>> which only have a GSC bus...
>>>>>
>>>>> We need another patch.
>>>>> Do you have a link to the build error?
>>>>
>>>>
>>>> No, it's from the other build error that you just replied to,
>>>> where the incorrect compiler was used.
>>>>
>>>> I'll recheck it and reconsider what to do, if anything.
>>>
>>> Ok, thank you!
>>
>> I dunno what to do. This:
>>
>> #ifdef CONFIG_64BIT
>> 	if (!dev->irq && (dev->id.sversion == 0xad))
>> 		dev->irq = iosapic_serial_irq(dev);
>> #endif
>>
>> makes it look like 64BIT requires IOSAPIC (hence PCI_LBA).
>
> Although I think all 64bit machines have a PCI bus, the better
> fix is that the driver should only call iosapic_serial_irq(dev)
> if CONFIG_IOSAPIC is set. This patch fixes the build:
>
> -#ifdef CONFIG_64BIT
> +#ifdef CONFIG_IOSAPIC
>         if (!dev->irq && (dev->id.sversion == 0xad))
>                 dev->irq = iosapic_serial_irq(dev);
>  #endif

That was not fully correct.
It needs to be:

#if defined(CONFIG_64BIT) && defined(CONFIG_IOSAPIC)

Otherwise you'll get an undefined reference in the 32-bit build.

Helge