linux-kernel - Re: [rfc v2 4/6] dma-direct: atomic allocations must come from atomic coherent pools

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 for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Tue, 14 Apr 2020 08:43:32 +0200
From:   Christoph Hellwig <hch@....de>
To:     David Rientjes <rientjes@...gle.com>
Cc:     Christoph Hellwig <hch@....de>,
        Tom Lendacky <thomas.lendacky@....com>,
        "Singh, Brijesh" <brijesh.singh@....com>,
        "Grimm, Jon" <jon.grimm@....com>, Joerg Roedel <joro@...tes.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "iommu@...ts.linux-foundation.org" <iommu@...ts.linux-foundation.org>
Subject: Re: [rfc v2 4/6] dma-direct: atomic allocations must come from
 atomic coherent pools

> +	/*
> +	 * Unencrypted memory must come directly from DMA atomic pools if
> +	 * blocking is not allowed.
> +	 */
> +	if (IS_ENABLED(CONFIG_DMA_COHERENT_POOL) &&
> +	    force_dma_unencrypted(dev) && !gfpflags_allow_blocking(gfp)) {
> +		ret = dma_alloc_from_pool(dev, PAGE_ALIGN(size), &page, gfp);
> +		if (!ret)
> +			return NULL;
> +		goto done;
> +	}
> +
>  	if (IS_ENABLED(CONFIG_DMA_DIRECT_REMAP) &&
>  	    dma_alloc_need_uncached(dev, attrs) &&
>  	    !gfpflags_allow_blocking(gfp)) {

Can we keep a single conditional for the pool allocations?  Maybe
add a new dma_alloc_from_pool helper ala:

static inline bool dma_alloc_from_pool(struct device *dev, gfp_t gfp)
{
	if (!IS_ENABLED(CONFIG_DMA_COHERENT_POOL))
		return false;
	if (gfpflags_allow_blocking(gfp))
		return false;
	if (force_dma_unencrypted(dev))
		return true;
	if (dma_alloc_need_uncached(dev))
		return true;
}
}

> @@ -203,6 +215,10 @@ void dma_direct_free_pages(struct device *dev, size_t size, void *cpu_addr,
>  {
>  	unsigned int page_order = get_order(size);
>  
> +	if (IS_ENABLED(CONFIG_DMA_COHERENT_POOL) &&
> +	    dma_free_from_pool(dev, cpu_addr, PAGE_ALIGN(size)))
> +		return;
> +

Similarly I think we should have a single conditional to free from the
pool instead.