lkml.org 
[lkml]   [2015]   [Apr]   [10]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Date
From
SubjectRe: [PATCH 2/2] soc: qcom: Add Shared Memory Manager driver
On Fri, Apr 03, 2015 at 04:03:20PM -0700, Bjorn Andersson wrote:
<snip>

> +static int qcom_smem_alloc_private(struct qcom_smem *smem,
> + unsigned host,
> + unsigned item,
> + size_t size)
> +{

<snip>

> + alloc_size = sizeof(*hdr) + ALIGN(size, 8);
> + if (p + alloc_size >= (void *)phdr + phdr->offset_free_uncached) {
> + dev_err(smem->dev, "Out of memory\n");
> + return -ENOSPC;
> + }

This check always fails due to the fact that we always get a ptr that points to
something beyond the free_uncached area. We ought to use:
alloc_size > phdr->offset_free_cached - phdr->offset_free_uncached

> +
> + hdr = p;
> + hdr->canary = SMEM_PRIVATE_CANARY;
> + hdr->item = item;
> + hdr->size = ALIGN(size, 8);
> + hdr->padding_data = hdr->size - size;
> + hdr->padding_hdr = 0;
> +

<snip>

> +static int qcom_smem_probe(struct platform_device *pdev)
> +{

<snip>

> + ret = of_address_to_resource(np, 0, &r);
> + of_node_put(np);
> + if (ret)
> + return ret;
> +
> + smem->regions[0].aux_base = (u32)r.start;
> + smem->regions[0].size = resource_size(&r);
> + smem->regions[0].virt_base = devm_ioremap(&pdev->dev,
> + r.start,
> + resource_size(&r));

Need to use devm_ioremap_nocache() instead. We need uncached accesses.

> + if (!smem->regions[0].virt_base)
> + return -ENOMEM;
> +
> + for (i = 1; i < num_regions; i++) {
> + res = platform_get_resource(pdev, IORESOURCE_MEM, i - 1);
> +
> + smem->regions[i].aux_base = (u32)res->start;
> + smem->regions[i].size = resource_size(res);
> + smem->regions[i].virt_base = devm_ioremap(&pdev->dev,
> + res->start,
> + resource_size(res));

Same thing here.

> + if (!smem->regions[i].virt_base)
> + return -ENOMEM;
> + }
> +

<snip>

> diff --git a/include/linux/soc/qcom/smem.h b/include/linux/soc/qcom/smem.h
> new file mode 100644
> index 0000000..294070de
> --- /dev/null
> +++ b/include/linux/soc/qcom/smem.h
> @@ -0,0 +1,14 @@
> +#ifndef __QCOM_SMEM_H__
> +#define __QCOM_SMEM_H__
> +
> +struct device_node;
> +struct qcom_smem;
> +
> +#define QCOM_SMEM_HOST_ANY -1
Would it make sense to throw in the remote processor enumeration? Same with the
fixed/dynamic item list?

> +
> +int qcom_smem_alloc(unsigned host, unsigned item, size_t size);
> +int qcom_smem_get(unsigned host, unsigned item, void **ptr, size_t *size);
> +
> +int qcom_smem_get_free_space(unsigned host);


--
Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
a Linux Foundation Collaborative Project



\
 
 \ /
  Last update: 2015-04-11 00:01    [W:0.068 / U:14.076 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site