Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Sat, 25 Apr 2020 02:03:11 +0800
From:   Jiaxun Yang <jiaxun.yang@flygoat.com>
To:     Rob Herring <robh+dt@kernel.org>
CC:     PCI <linux-pci@vger.kernel.org>,
        Bjorn Helgaas <bhelgaas@google.com>,
        Thomas Bogendoerfer <tsbogend@alpha.franken.de>,
        Huacai Chen <chenhc@lemote.com>,
        Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>,
        Andrew Murray <amurray@thegoodpenguin.co.uk>,
        Paul Burton <paulburton@kernel.org>,
        devicetree@vger.kernel.org,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        "open list:MIPS" <linux-mips@vger.kernel.org>
Subject: Re: [PATCH v5 1/6] PCI: OF: Don't remap iospace on unsupported platform
User-Agent: K-9 Mail for Android
Reply-to: jiaxun.yang@flygoat.com
In-Reply-To: <CAL_JsqKsZrsnn92XHL4QdFkAgFKqVTBZXDz9wVv-_7oywOOsBQ@mail.gmail.com>
References: <20200424130847.328584-1-jiaxun.yang@flygoat.com> <20200424130847.328584-2-jiaxun.yang@flygoat.com> <CAL_JsqKsZrsnn92XHL4QdFkAgFKqVTBZXDz9wVv-_7oywOOsBQ@mail.gmail.com>
Message-ID: <D5CC0902-F84F-43F1-88E5-C03018B25050@flygoat.com>
MIME-Version: 1.0
Content-Type: text/plain;
 charset=utf-8
Content-Transfer-Encoding: quoted-printable
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk


=E4=BA=8E 2020=E5=B9=B44=E6=9C=8825=E6=97=A5 GMT+08:00 =E4=B8=8A=E5=8D=881=
:47:23, Rob Herring <robh+dt@kernel=2Eorg> =E5=86=99=E5=88=B0:
>On Fri, Apr 24, 2020 at 8:09 AM Jiaxun Yang <jiaxun=2Eyang@flygoat=2Ecom>=
 wrote:
>>
>> There are some platforms that don't support I/O space remapping
>> like MIPS=2E However, our PCI code will try to remap iospace
>> unconditionally and reject io resources on these platforms=2E
>>
>> So we should remove I/O space remapping check and use a range
>> check instead on these platforms=2E
>>
>> Signed-off-by: Jiaxun Yang <jiaxun=2Eyang@flygoat=2Ecom>
>> --
>> v4: Fix a typo in commit message=2E
>> v5: Commit message massage
>> ---
>>  drivers/pci/of=2Ec | 9 +++++++++
>>  1 file changed, 9 insertions(+)
>>
>> diff --git a/drivers/pci/of=2Ec b/drivers/pci/of=2Ec
>> index 81ceeaa6f1d5=2E=2E36e8761b66c6 100644
>> --- a/drivers/pci/of=2Ec
>> +++ b/drivers/pci/of=2Ec
>> @@ -547,12 +547,21 @@ int pci_parse_request_of_pci_ranges(struct device=
 *dev,
>>
>>                 switch (resource_type(res)) {
>>                 case IORESOURCE_IO:
>> +#if defined(PCI_IOBASE) && defined(CONFIG_MMU)
>
>We already have the same condition in pci_remap_iospace()=2E All this
>does is suppress a WARN=2E If a WARN is not desired, then change it=2E
>Perhaps just a single line rather than backtrace would be okay=2E
>Whether to WARN or not shouldn't be a decision for firmware code=2E
>
>Though really, if I/O space can never be supported, then it's an error
>in your DT to describe it=2E

In MIPS world, we do have inb/oub family of I/O functions
to support I/O space=2E But we're not using remap_iospace as it's breaking
some of our assumptions=2E
We have our own inb/outb implementation=2E

In that case, I think make a range check instead of remapping would
be more practical=2E

Thanks=2E

>
>>                         err =3D devm_pci_remap_iospace(dev, res, iobase=
);
>>                         if (err) {
>>                                 dev_warn(dev, "error %d: failed to map =
resource %pR\n",
>>                                          err, res);
>>                                 resource_list_destroy_entry(win);
>>                         }
>> +#else
>> +                       /* Simply check if IO is inside the range */
>
>Why do you care if it's never used?
>
>> +                       if (res->end > IO_SPACE_LIMIT) {
>> +                               dev_warn(dev, "resource %pR out of the =
I/O range\n",
>> +                                       res);
>> +                               resource_list_destroy_entry(win);
>> +                       }
>> +#endif
>>                         break;
>>                 case IORESOURCE_MEM:
>>                         res_valid |=3D !(res->flags & IORESOURCE_PREFET=
CH);
>> --
>> 2=2E26=2E0=2Erc2
>>

--=20
Jiaxun Yang