From mboxrd@z Thu Jan  1 00:00:00 1970
From: Florian Fainelli <f.fainelli@gmail.com>
Subject: Re: [PATCH v3 4/6] arm64: Utilize phys_initrd_start/phys_initrd_size
Date: Mon, 5 Nov 2018 12:51:23 -0800
Message-ID: <0257ee07-8b85-5295-1490-72c7aa14943d@gmail.com>
References: <20181031192843.13230-1-f.fainelli@gmail.com>
 <20181031192843.13230-5-f.fainelli@gmail.com>
 <CAKv+Gu85kRVos-z+1WuV5J8bTHpLDDdUu0hVD2NmZ-w-ry5f4g@mail.gmail.com>
 <537b8384-629a-770b-e506-f4d49b92e758@gmail.com>
 <CAKv+Gu88HHe-ZpVS3w+0Eu+Q-zYaN3sKDK4uaG+eecVtr-C9PA@mail.gmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8bit
Return-path: <linux-kernel-owner@vger.kernel.org>
In-Reply-To: <CAKv+Gu88HHe-ZpVS3w+0Eu+Q-zYaN3sKDK4uaG+eecVtr-C9PA@mail.gmail.com>
Content-Language: en-US
Sender: linux-kernel-owner@vger.kernel.org
To: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Linux Kernel Mailing List <linux-kernel@vger.kernel.org>, Catalin Marinas <catalin.marinas@arm.com>, Will Deacon <will.deacon@arm.com>, Rob Herring <robh+dt@kernel.org>, Frank Rowand <frowand.list@gmail.com>, Andrew Morton <akpm@linux-foundation.org>, Marc Zyngier <marc.zyngier@arm.com>, Russell King <rmk+kernel@armlinux.org.uk>, Andrey Ryabinin <aryabinin@virtuozzo.com>, Andrey Konovalov <andreyknvl@google.com>, Masahiro Yamada <yamada.masahiro@socionext.com>, Robin Murphy <robin.murphy@arm.com>, Laura Abbott <labbott@redhat.com>, Stefan Agner <stefan@agner.ch>, Johannes Weiner <hannes@cmpxchg.org>, Greg Hackmann <ghackmann@android.com>, Kristina Martsenko <kristina.martsenko@arm.com>, CHANDAN VN <chandan.vn@samsung.com>, moderated list:ARM
List-Id: devicetree@vger.kernel.org

On 11/5/18 12:44 PM, Ard Biesheuvel wrote:
> On 5 November 2018 at 21:41, Florian Fainelli <f.fainelli@gmail.com> wrote:
>> On 11/5/18 12:39 PM, Ard Biesheuvel wrote:
>>> Hi Florian,
>>>
>>> On 31 October 2018 at 20:28, Florian Fainelli <f.fainelli@gmail.com> wrote:
>>>> ARM64 is the only architecture that re-defines
>>>> __early_init_dt_declare_initrd() in order for that function to populate
>>>> initrd_start/initrd_end with physical addresses instead of virtual
>>>> addresses. Instead of having an override we can leverage
>>>> drivers/of/fdt.c populating phys_initrd_start/phys_initrd_size to
>>>> populate those variables for us.
>>>>
>>>> Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
>>>> ---
>>>>  arch/arm64/mm/init.c | 19 +++++++++----------
>>>>  1 file changed, 9 insertions(+), 10 deletions(-)
>>>>
>>>> diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c
>>>> index 3cf87341859f..00ef2166bb73 100644
>>>> --- a/arch/arm64/mm/init.c
>>>> +++ b/arch/arm64/mm/init.c
>>>> @@ -72,8 +72,8 @@ static int __init early_initrd(char *p)
>>>>         if (*endp == ',') {
>>>>                 size = memparse(endp + 1, NULL);
>>>>
>>>> -               initrd_start = start;
>>>> -               initrd_end = start + size;
>>>> +               phys_initrd_start = start;
>>>> +               phys_initrd_size = size;
>>>>         }
>>>>         return 0;
>>>>  }
>>>> @@ -408,14 +408,14 @@ void __init arm64_memblock_init(void)
>>>>                 memblock_add(__pa_symbol(_text), (u64)(_end - _text));
>>>>         }
>>>>
>>>> -       if (IS_ENABLED(CONFIG_BLK_DEV_INITRD) && initrd_start) {
>>>> +       if (IS_ENABLED(CONFIG_BLK_DEV_INITRD) && phys_initrd_size) {
>>>>                 /*
>>>>                  * Add back the memory we just removed if it results in the
>>>>                  * initrd to become inaccessible via the linear mapping.
>>>>                  * Otherwise, this is a no-op
>>>>                  */
>>>> -               u64 base = initrd_start & PAGE_MASK;
>>>> -               u64 size = PAGE_ALIGN(initrd_end) - base;
>>>> +               u64 base = phys_initrd_start & PAGE_MASK;
>>>> +               u64 size = PAGE_ALIGN(phys_initrd_size);
>>>>
>>>>                 /*
>>>>                  * We can only add back the initrd memory if we don't end up
>>>> @@ -460,12 +460,11 @@ void __init arm64_memblock_init(void)
>>>>          */
>>>>         memblock_reserve(__pa_symbol(_text), _end - _text);
>>>>  #ifdef CONFIG_BLK_DEV_INITRD
>>>> -       if (initrd_start) {
>>>> -               memblock_reserve(initrd_start, initrd_end - initrd_start);
>>>> -
>>>> +       if (phys_initrd_size) {
>>>>                 /* the generic initrd code expects virtual addresses */
>>>> -               initrd_start = __phys_to_virt(initrd_start);
>>>> -               initrd_end = __phys_to_virt(initrd_end);
>>>> +               initrd_start = __phys_to_virt(phys_initrd_start);
>>>> +               initrd_end = initrd_start + phys_initrd_size;
>>>> +               initrd_below_start_ok = 0;
>>>
>>> Where is this assignment coming from?
>>
>> __early_init_dt_declare_initrd() sets initrd_below_start_ok to 1 though
>> after patch #5 this is not necessary any more.
> 
> Yes, but why? The original arm64 version of
> __early_init_dt_declare_initrd() does not set it but now you set to 1
> in the IS_ENABLED(CONFIG_ARM64) section in the generic code and set it
> back to 0 here.

Humm, it is an if (!IS_ENABLED(CONFIG_ARM64)) condition, so we would not
be taking that branch on an ARM64 kernel.

If you are saying the assignment is not necessary anymore after patch #5
, that is true, though this can only be done a part of part #5, not as
part of patch #4 in order not to break initrd functionality in-between
patches.

> 
> Or am I missing something?
> 

Not sure, I could be too, it's Monday after all :)
-- 
Florian