Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Fri, 29 Apr 2022 08:28:34 -0700 (PDT)
Subject:     Re: [PATCH] RISC-V-fixes: relocate DTB if it's outside memory region
In-Reply-To: <mhng-9aafbd90-eb33-4f43-afa3-b9f9338c2a70@palmer-ri-x1c9>
CC:     Paul Walmsley <paul.walmsley@sifive.com>, aou@eecs.berkeley.edu,
        linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org
From:   Palmer Dabbelt <palmer@dabbelt.com>
To:     mick@ics.forth.gr
Message-ID: <mhng-6c123e84-9530-46f6-a5ec-6b908bb92f95@palmer-ri-x1c9>
Mime-Version: 1.0 (MHng)
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk

On Thu, 28 Apr 2022 14:48:14 PDT (-0700), Palmer Dabbelt wrote:
> On Mon, 25 Apr 2022 23:11:23 PDT (-0700), mick@ics.forth.gr wrote:
>> Hello Palmer,
>>
>> Any updates on this ?
>
> Sorry about that, it's on fixes.

Not sure if I just wasn't paying attention yesterday or if I'm grumpier 
this morning, but that "RISC-V-fixes: " prefix is just a bit too odd -- 
I know we've got a split between "RISC-V" and "riscv" so maybe it 
doesn't matter, but even that is kind of ugly.

I re-wrote it, but I'm going to let it round trip through linux-next so 
I'll send it up next time.

Sorry, I know this happened twice recently but I'll try not to make a 
habit of it.

>
>>
>> Regards,
>> Nick
>>
>> On 3/22/22 15:28, Nick Kossifidis wrote:
>>> In case the DTB provided by the bootloader/BootROM is before the kernel
>>> image or outside /memory, we won't be able to access it through the
>>> linear mapping, and get a segfault on setup_arch(). Currently OpenSBI
>>> relocates DTB but that's not always the case (e.g. if FW_JUMP_FDT_ADDR
>>> is not specified), and it's also not the most portable approach since
>>> the default FW_JUMP_FDT_ADDR of the generic platform relocates the DTB
>>> at a specific offset that may not be available. To avoid this situation
>>> copy DTB so that it's visible through the linear mapping.
>>>
>>> Signed-off-by: Nick Kossifidis <mick@ics.forth.gr>
>>> ---
>>>   arch/riscv/mm/init.c | 21 +++++++++++++++++++--
>>>   1 file changed, 19 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/arch/riscv/mm/init.c b/arch/riscv/mm/init.c
>>> index 0d588032d..697a9aed4 100644
>>> --- a/arch/riscv/mm/init.c
>>> +++ b/arch/riscv/mm/init.c
>>> @@ -206,8 +206,25 @@ static void __init setup_bootmem(void)
>>>   	 * early_init_fdt_reserve_self() since __pa() does
>>>   	 * not work for DTB pointers that are fixmap addresses
>>>   	 */
>>> -	if (!IS_ENABLED(CONFIG_BUILTIN_DTB))
>>> -		memblock_reserve(dtb_early_pa, fdt_totalsize(dtb_early_va));
>>> +	if (!IS_ENABLED(CONFIG_BUILTIN_DTB)) {
>>> +		/*
>>> +		 * In case the DTB is not located in a memory region we won't
>>> +		 * be able to locate it later on via the linear mapping and
>>> +		 * get a segfault when accessing it via __va(dtb_early_pa).
>>> +		 * To avoid this situation copy DTB to a memory region.
>>> +		 * Note that memblock_phys_alloc will also reserve DTB region.
>>> +		 */
>>> +		if (!memblock_is_memory(dtb_early_pa)) {
>>> +			size_t fdt_size = fdt_totalsize(dtb_early_va);
>>> +			phys_addr_t new_dtb_early_pa = memblock_phys_alloc(fdt_size, PAGE_SIZE);
>>> +			void *new_dtb_early_va = early_memremap(new_dtb_early_pa, fdt_size);
>>> +
>>> +			memcpy(new_dtb_early_va, dtb_early_va, fdt_size);
>>> +			early_memunmap(new_dtb_early_va, fdt_size);
>>> +			_dtb_early_pa = new_dtb_early_pa;
>>> +		} else
>>> +			memblock_reserve(dtb_early_pa, fdt_totalsize(dtb_early_va));
>>> +	}
>>>
>>>   	early_init_fdt_scan_reserved_mem();
>>>   	dma_contiguous_reserve(dma32_phys_limit);