Re: riscv32 EXT4 splat, 6.8 regression?

From: Mike Rapoport <rppt@kernel.org>
To: Nam Cao <namcao@linutronix.de>
Cc: "Björn Töpel" <bjorn@kernel.org>,
	"Christian Brauner" <brauner@kernel.org>,
	"Andreas Dilger" <adilger@dilger.ca>,
	"Al Viro" <viro@zeniv.linux.org.uk>,
	linux-fsdevel <linux-fsdevel@vger.kernel.org>,
	"Jan Kara" <jack@suse.cz>,
	"Linux Kernel Mailing List" <linux-kernel@vger.kernel.org>,
	linux-riscv@lists.infradead.org, "Theodore Ts'o" <tytso@mit.edu>,
	"Ext4 Developers List" <linux-ext4@vger.kernel.org>,
	"Conor Dooley" <conor@kernel.org>,
	"Matthew Wilcox (Oracle)" <willy@infradead.org>,
	"Anders Roxell" <anders.roxell@linaro.org>,
	"Alexandre Ghiti" <alex@ghiti.fr>
Subject: Re: riscv32 EXT4 splat, 6.8 regression?
Date: Tue, 16 Apr 2024 19:31:54 +0300	[thread overview]
Message-ID: <Zh6n-nvnQbL-0xss@kernel.org> (raw)
In-Reply-To: <20240416181944.23af44ee@namcao>

On Tue, Apr 16, 2024 at 06:19:44PM +0200, Nam Cao wrote:
> On 2024-04-16 Björn Töpel wrote:
> > Nam Cao <namcao@linutronix.de> writes:
> > 
> > > Fixed version:
> > >
> > > diff --git a/arch/riscv/mm/init.c b/arch/riscv/mm/init.c
> > > index fa34cf55037b..af4192bc51d0 100644
> > > --- a/arch/riscv/mm/init.c
> > > +++ b/arch/riscv/mm/init.c
> > > @@ -245,6 +245,7 @@ static void __init setup_bootmem(void)
> > >  	 * be done as soon as the kernel mapping base address is determined.
> > >  	 */
> > >  	if (!IS_ENABLED(CONFIG_64BIT)) {
> > > +		memblock_reserve(__pa(-PAGE_SIZE), PAGE_SIZE);
> > >  		max_mapped_addr = __pa(~(ulong)0);
> > >  		if (max_mapped_addr == (phys_ram_end - 1))
> > >  			memblock_set_current_limit(max_mapped_addr - 4096);  
> > 
> > Nice!
> > 
> > Can't we get rid of the if-statement, and max_mapped_address as well?
> 
> I don't see why not :D
> 
> Best regards,
> Nam
> 
> diff --git a/arch/riscv/mm/init.c b/arch/riscv/mm/init.c
> index fa34cf55037b..f600cfee0aef 100644
> --- a/arch/riscv/mm/init.c
> +++ b/arch/riscv/mm/init.c
> @@ -197,7 +197,6 @@ early_param("mem", early_mem);
>  static void __init setup_bootmem(void)
>  {
>  	phys_addr_t vmlinux_end = __pa_symbol(&_end);
> -	phys_addr_t max_mapped_addr;
>  	phys_addr_t phys_ram_end, vmlinux_start;
>  
>  	if (IS_ENABLED(CONFIG_XIP_KERNEL))
> @@ -238,17 +237,9 @@ static void __init setup_bootmem(void)
>  	/*
>  	 * memblock allocator is not aware of the fact that last 4K bytes of
>  	 * the addressable memory can not be mapped because of IS_ERR_VALUE
> -	 * macro. Make sure that last 4k bytes are not usable by memblock
> -	 * if end of dram is equal to maximum addressable memory.  For 64-bit
> -	 * kernel, this problem can't happen here as the end of the virtual
> -	 * address space is occupied by the kernel mapping then this check must
> -	 * be done as soon as the kernel mapping base address is determined.
> +	 * macro. Make sure that last 4k bytes are not usable by memblock.
>  	 */

It's not only memblock, but buddy as well, so maybe

	/*
	 * The last 4K bytes of the addressable memory can not be used
	 * because of IS_ERR_VALUE macro. Make sure that last 4K bytes are
	 * not usable by kernel memory allocators.
	 */

> -	if (!IS_ENABLED(CONFIG_64BIT)) {
> -		max_mapped_addr = __pa(~(ulong)0);
> -		if (max_mapped_addr == (phys_ram_end - 1))
> -			memblock_set_current_limit(max_mapped_addr - 4096);
> -	}
> +	memblock_reserve(__pa(-PAGE_SIZE), PAGE_SIZE);

Ack.

>  	min_low_pfn = PFN_UP(phys_ram_base);
>  	max_low_pfn = max_pfn = PFN_DOWN(phys_ram_end);
> 

-- 
Sincerely yours,
Mike.