linux-kernel - Re: [PATCH v5 2/2] iommu/mediatek: Allow page table PA up to 35bit

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] [day] [month] [year] [list]

Date:   Sat, 11 Jun 2022 18:20:38 +0800
From:   <yf.wang@...iatek.com>
To:     <yong.wu@...iatek.com>
CC:     <Libo.Kang@...iatek.com>, <iommu@...ts.linux-foundation.org>,
        <joro@...tes.org>, <linux-arm-kernel@...ts.infradead.org>,
        <linux-kernel@...r.kernel.org>,
        <linux-mediatek@...ts.infradead.org>, <matthias.bgg@...il.com>,
        <ning.li@...iatek.com>, <robin.murphy@....com>, <will@...nel.org>,
        <miles.chen@...iatek.com>, <wsd_upstream@...iatek.com>,
        <yf.wang@...iatek.com>
Subject: Re: [PATCH v5 2/2] iommu/mediatek: Allow page table PA up to 35bit

On Thu, 2022-05-19 at 14:58 +0800, Yong Wu wrote:
> On Mon, 2022-05-16 at 22:16 +0800, yf.wang@...iatek.com wrote:
> > From: Yunfei Wang <yf.wang@...iatek.com>
> > 
> > Add the quirk IO_PGTABLE_QUIRK_ARM_MTK_TTBR_EXT support, so that
> > allows
> > page table PA up to 35bit, not only in ZONE_DMA32.
> 
> Comment why this is needed.
> 
> e.g. For single normal zone.
> 


Hi Yong,

There is no DMA32 zone in some mediatek smartphone chip for single
normal zone.

The level 1 and level 2 pgtable are both allocated in ZONE_DMA32,
and may have two possible problem:
1.The level 2 pgtable is allocated in ZONE_DMA32 with atomic flag,
 and it may fail if ZONE_DMA32 memory is used out.
2.Single memory feature will make ZONE_DMA32 empty, and cause level
 1 and level 2 pgtable PA more than 32bit.

Solution:
Add the quirk IO_PGTABLE_QUIRK_ARM_MTK_TTBR_EXT support, so that
level 1 and level 2 pgtable can support at most 35bit PA.

I had updated the commit message in V7 version.


> >  		data->m4u_dom = dom;
> > -		writel(dom->cfg.arm_v7s_cfg.ttbr & MMU_PT_ADDR_MASK,
> > -		       data->base + REG_MMU_PT_BASE_ADDR);
> > +
> > +		/* Bits[6:3] are invalid for mediatek platform */
> > +		if (MTK_IOMMU_HAS_FLAG(data->plat_data,
> > PGTABLE_PA_35_EN))
> > +			regval = (dom->cfg.arm_v7s_cfg.ttbr &
> > MMU_PT_ADDR_MASK) |
> > +				 (dom->cfg.arm_v7s_cfg.ttbr &
> > MMU_PT_ADDR_2_0_MASK);
> 
> The bits[2:0] has already handled in ARM_V7S_TTBR_35BIT_PA of
> patch[1/2], we need calculate it again here?
> 
> 1) Extend arm_v7s_cfg.ttbr to u64, then we could put the special ttbr
> logical into our file.
> 
> 2) if 1) is not allowed, We have to put this in v7s.
> if (cfg->quirks & IO_PGTABLE_QUIRK_ARM_MTK_TTBR_EXT) {
>        cfg->arm_v7s_cfg.ttbr = (paddr & GENMASK(31, 7)) |
> upper_32_bits(paddr);
>        return &data->iop;
> }
> 
> This need Robin/Will confirm.
> 


Hi Yong,

V7 version had extended arm_v7s_cfg.ttbr to u64, then put the special
ttbr logical into MediaTek file.


> > +		else
> > +			regval = dom->cfg.arm_v7s_cfg.ttbr &
> > MMU_PT_ADDR_MASK;
> 
> Save this value to our data. then we don't need calculate it again in
> the resume cb.   
> 

Hi Yong,

V8 version will Save the special ttbr to mtk_iommu_domain avoid
calculate it again.

Thanks,
Yunfei.