Message-ID: <4vwahy5pb74an2nebp643s2doacbrslrsitolzksmmjnv4lxxu@vub4gpex5fcs>
Date: Fri, 7 Jun 2024 14:41:46 +0300
From: Dmitry Baryshkov <dmitry.baryshkov@...aro.org>
To: Ekansh Gupta <quic_ekangupt@...cinc.com>
Cc: srinivas.kandagatla@...aro.org, linux-arm-msm@...r.kernel.org, 
	gregkh@...uxfoundation.org, quic_bkumar@...cinc.com, linux-kernel@...r.kernel.org, 
	quic_chennak@...cinc.com, stable <stable@...nel.org>
Subject: Re: [PATCH v4 09/11] misc: fastrpc: Fix remote heap alloc and free
 user request

On Thu, Jun 06, 2024 at 10:29:29PM +0530, Ekansh Gupta wrote:
> A static PD daemon process can request for remote heap allocation
> which will allocate memory and change the ownership if the VMID
> information were defined. This allocations are currently getting
> added to fl mmaps list which could get freed when there is any
> daemon kill. There is no way to request for freeing of this memory
> also. Add changes to maintain the remote heap allocation in the

Simpler. 'Maintain foo'

> static PD specific structure and add method to free the memory
> on user request.

Should this be split into two patches? 'foo and bar' suggests that.

> 
> Fixes: 532ad70c6d44 ("misc: fastrpc: Add mmap request assigning for static PD pool")
> Cc: stable <stable@...nel.org>
> Signed-off-by: Ekansh Gupta <quic_ekangupt@...cinc.com>
> ---
>  drivers/misc/fastrpc.c | 129 +++++++++++++++++++++++++++++++----------
>  1 file changed, 98 insertions(+), 31 deletions(-)
> 
> diff --git a/drivers/misc/fastrpc.c b/drivers/misc/fastrpc.c
> index 68c1595446d5..32f2e6f625ed 100644
> --- a/drivers/misc/fastrpc.c
> +++ b/drivers/misc/fastrpc.c
> @@ -210,6 +210,7 @@ struct fastrpc_buf {
>  	struct dma_buf *dmabuf;
>  	struct device *dev;
>  	void *virt;
> +	u32 flag;
>  	u64 phys;
>  	u64 size;
>  	/* Lock for dma buf attachments */
> @@ -1924,29 +1925,54 @@ static int fastrpc_get_dsp_info(struct fastrpc_user *fl, char __user *argp)
>  	return 0;
>  }
>  
> -static int fastrpc_req_munmap_impl(struct fastrpc_user *fl, struct fastrpc_buf *buf)
> +static int fastrpc_req_munmap_dsp(struct fastrpc_user *fl, uintptr_t raddr, u64 size)
>  {
>  	struct fastrpc_invoke_args args[1] = { [0] = { 0 } };
>  	struct fastrpc_munmap_req_msg req_msg;
> -	struct device *dev = fl->sctx->dev;
>  	int err;
>  	u32 sc;
>  
>  	req_msg.pgid = fl->tgid;
> -	req_msg.size = buf->size;
> -	req_msg.vaddr = buf->raddr;
> +	req_msg.size = size;
> +	req_msg.vaddr = raddr;
>  
>  	args[0].ptr = (u64) (uintptr_t) &req_msg;
>  	args[0].length = sizeof(req_msg);
> -
>  	sc = FASTRPC_SCALARS(FASTRPC_RMID_INIT_MUNMAP, 1, 0);
>  	err = fastrpc_internal_invoke(fl, true, FASTRPC_INIT_HANDLE, sc,
>  				      &args[0]);
> +
> +	return err;
> +}
> +
> +static int fastrpc_req_munmap_impl(struct fastrpc_user *fl, struct fastrpc_buf *buf)
> +{
> +	struct device *dev = fl->sctx->dev;
> +	int err;
> +
> +	err = fastrpc_req_munmap_dsp(fl, buf->raddr, buf->size);
>  	if (!err) {
> +		if (buf->flag == ADSP_MMAP_REMOTE_HEAP_ADDR) {
> +			if (fl->cctx->vmcount) {
> +				u64 src_perms = 0;
> +				struct qcom_scm_vmperm dst_perms;
> +				u32 i;
> +
> +				for (i = 0; i < fl->cctx->vmcount; i++)
> +					src_perms |= BIT(fl->cctx->vmperms[i].vmid);
> +
> +				dst_perms.vmid = QCOM_SCM_VMID_HLOS;
> +				dst_perms.perm = QCOM_SCM_PERM_RWX;
> +				err = qcom_scm_assign_mem(buf->phys, (u64)buf->size,
> +								&src_perms, &dst_perms, 1);
> +				if (err) {
> +					dev_err(dev, "Failed to assign memory phys 0x%llx size 0x%llx err %d\n",
> +						buf->phys, buf->size, err);
> +					return err;
> +				}

It looks like this is too nested. Consider refactoring this code. For
example, extract the function that you have c&p'ed.

> +			}
> +		}
>  		dev_dbg(dev, "unmmap\tpt 0x%09lx OK\n", buf->raddr);
> -		spin_lock(&fl->lock);
> -		list_del(&buf->node);
> -		spin_unlock(&fl->lock);
>  		fastrpc_buf_free(buf);
>  	} else {
>  		dev_err(dev, "unmmap\tpt 0x%09lx ERROR\n", buf->raddr);
> @@ -1960,6 +1986,7 @@ static int fastrpc_req_munmap(struct fastrpc_user *fl, char __user *argp)
>  	struct fastrpc_buf *buf = NULL, *iter, *b;
>  	struct fastrpc_req_munmap req;
>  	struct device *dev = fl->sctx->dev;
> +	int err = 0;

Why =0 is necessary?

>  
>  	if (copy_from_user(&req, argp, sizeof(req)))
>  		return -EFAULT;
> @@ -1968,18 +1995,45 @@ static int fastrpc_req_munmap(struct fastrpc_user *fl, char __user *argp)
>  	list_for_each_entry_safe(iter, b, &fl->mmaps, node) {
>  		if ((iter->raddr == req.vaddrout) && (iter->size == req.size)) {
>  			buf = iter;
> +			list_del(&buf->node);
>  			break;
>  		}
>  	}
>  	spin_unlock(&fl->lock);
>  
> -	if (!buf) {
> -		dev_err(dev, "mmap\t\tpt 0x%09llx [len 0x%08llx] not in list\n",
> -			req.vaddrout, req.size);
> -		return -EINVAL;
> +	if (buf) {
> +		err = fastrpc_req_munmap_impl(fl, buf);
> +		if (err) {
> +			spin_lock(&fl->lock);
> +			list_add_tail(&buf->node, &fl->mmaps);
> +			spin_unlock(&fl->lock);
> +		}
> +		return err;
>  	}
>  
> -	return fastrpc_req_munmap_impl(fl, buf);
> +	spin_lock(&fl->lock);
> +	if (fl->spd) {
> +		list_for_each_entry_safe(iter, b, &fl->spd->rmaps, node) {
> +			if ((iter->raddr == req.vaddrout) && (iter->size == req.size)) {
> +				buf = iter;
> +				list_del(&buf->node);
> +				break;
> +			}
> +		}
> +	}
> +	spin_unlock(&fl->lock);
> +	if (buf) {
> +		err = fastrpc_req_munmap_impl(fl, buf);
> +		if (err) {
> +			spin_lock(&fl->lock);
> +			list_add_tail(&buf->node, &fl->spd->rmaps);
> +			spin_unlock(&fl->lock);
> +		}
> +		return err;
> +	}
> +	dev_err(dev, "buffer not found addr 0x%09lx, len 0x%08llx\n",
> +			req.vaddrout, req.size);

Can this be triggered by the user? If so, it's dev_dbg() at best.

> +	return -EINVAL;
>  }
>  
>  static int fastrpc_req_mmap(struct fastrpc_user *fl, char __user *argp)
> @@ -2008,15 +2062,34 @@ static int fastrpc_req_mmap(struct fastrpc_user *fl, char __user *argp)
>  		return -EINVAL;
>  	}
>  
> -	if (req.flags == ADSP_MMAP_REMOTE_HEAP_ADDR)
> +	if (req.flags == ADSP_MMAP_REMOTE_HEAP_ADDR) {
> +		if (!fl->spd || !fl->spd->ispdup) {
> +			dev_err(dev, "remote heap request supported only for active static PD\n");
> +			return -EINVAL;
> +		}
>  		err = fastrpc_remote_heap_alloc(fl, dev, req.size, &buf);
> -	else
> +	} else {
>  		err = fastrpc_buf_alloc(fl, dev, req.size, &buf);
> +	}
>  
>  	if (err) {
>  		dev_err(dev, "failed to allocate buffer\n");
>  		return err;
>  	}
> +	buf->flag = req.flags;
> +
> +	/* Add memory to static PD pool, protection through hypervisor */
> +	if ((req.flags == ADSP_MMAP_REMOTE_HEAP_ADDR) && fl->cctx->vmcount) {
> +		u64 src_perms = BIT(QCOM_SCM_VMID_HLOS);
> +
> +		err = qcom_scm_assign_mem(buf->phys, (u64)buf->size,
> +			&src_perms, fl->cctx->vmperms, fl->cctx->vmcount);
> +		if (err) {
> +			dev_err(fl->sctx->dev, "Failed to assign memory phys 0x%llx size 0x%llx err %d\n",
> +					buf->phys, buf->size, err);

misaligned

> +			goto err_invoke;
> +		}
> +	}
>  
>  	req_msg.pgid = fl->tgid;
>  	req_msg.flags = req.flags;
> @@ -2049,26 +2122,16 @@ static int fastrpc_req_mmap(struct fastrpc_user *fl, char __user *argp)
>  	/* let the client know the address to use */
>  	req.vaddrout = rsp_msg.vaddr;
>  
> -	/* Add memory to static PD pool, protection thru hypervisor */
> -	if (req.flags == ADSP_MMAP_REMOTE_HEAP_ADDR && fl->cctx->vmcount) {
> -		u64 src_perms = BIT(QCOM_SCM_VMID_HLOS);
> -
> -		err = qcom_scm_assign_mem(buf->phys, (u64)buf->size,
> -			&src_perms, fl->cctx->vmperms, fl->cctx->vmcount);
> -		if (err) {
> -			dev_err(fl->sctx->dev, "Failed to assign memory phys 0x%llx size 0x%llx err %d",
> -					buf->phys, buf->size, err);
> -			goto err_assign;
> -		}
> -	}
> -
>  	spin_lock(&fl->lock);
> -	list_add_tail(&buf->node, &fl->mmaps);
> +	if (req.flags == ADSP_MMAP_REMOTE_HEAP_ADDR)
> +		list_add_tail(&buf->node, &fl->spd->rmaps);
> +	else
> +		list_add_tail(&buf->node, &fl->mmaps);
>  	spin_unlock(&fl->lock);
>  
>  	if (copy_to_user((void __user *)argp, &req, sizeof(req))) {
>  		err = -EFAULT;
> -		goto err_assign;
> +		goto err_copy;
>  	}
>  
>  	dev_dbg(dev, "mmap\t\tpt 0x%09lx OK [len 0x%08llx]\n",
> @@ -2076,8 +2139,12 @@ static int fastrpc_req_mmap(struct fastrpc_user *fl, char __user *argp)
>  
>  	return 0;
>  
> -err_assign:
> +err_copy:
> +	spin_lock(&fl->lock);
> +	list_del(&buf->node);
> +	spin_unlock(&fl->lock);
>  	fastrpc_req_munmap_impl(fl, buf);
> +	buf = NULL;
>  err_invoke:
>  	fastrpc_buf_free(buf);
>  
> -- 
> 2.43.0
> 

-- 
With best wishes
Dmitry

Powered by blists - more mailing lists