| 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
| ||
|
Message-ID: <20150816210725.GB17402@zurbaran.home>
Date: Sun, 16 Aug 2015 23:07:25 +0200
From: Samuel Ortiz <sameo@...ux.intel.com>
To: Robert Baldyga <r.baldyga@...sung.com>
Cc: lauro.venancio@...nbossa.org, aloisio.almeida@...nbossa.org,
robh+dt@...nel.org, pawel.moll@....com, mark.rutland@....com,
ijc+devicetree@...lion.org.uk, galak@...eaurora.org,
linux-wireless@...r.kernel.org, devicetree@...r.kernel.org,
linux-kernel@...r.kernel.org, linux-nfc@...ts.01.org,
m.szyprowski@...sung.com, pebolle@...cali.nl
Subject: Re: [PATCH v2] nfc: s3fwrn5: Add driver for Samsung S3FWRN5 NFC Chip
Hi Robert,
On Thu, Jul 30, 2015 at 04:26:16PM +0200, Robert Baldyga wrote:
> +static int s3fwrn5_firmware_update(struct s3fwrn5_info *info)
> +{
> + u32 version;
> + bool need_update;
> + int ret;
> +
> + ENTER();
We have many tracing tools and frameworks in the kernel, I don't think
this is needed.
> +static int s3fwrn5_nci_send(struct nci_dev *ndev, struct sk_buff *skb)
> +{
> + struct s3fwrn5_info *info = nci_get_drvdata(ndev);
> + int ret;
> +
> + ENTER();
> +
> + mutex_lock(&info->mutex);
> +
> + if (s3fwrn5_get_mode(info) != S3FWRN5_MODE_NCI)
Why can't we use this routine in firmware mode ?
> + return -EINVAL;
You must unlock before returning.
> +int s3fwrn5_recv_frame(struct nci_dev *ndev, struct sk_buff *skb,
> + enum s3fwrn5_mode mode)
> +{
> + struct s3fwrn5_info *info = nci_get_drvdata(ndev);
> +
> + switch (mode) {
> + case S3FWRN5_MODE_NCI:
> + return info->custom_nci ?
> + s3fwrn5_nci_recv_frame(ndev, skb) :
> + nci_recv_frame(ndev, skb);
In NCI mode, you should not have any of this logic implemented here, but
rely on the NCI core one instead.
> +int s3fwrn5_fw_get_base_addr(struct s3fwrn5_fw_cmd_get_bootinfo_rsp *bootinfo,
> + u32 *base_addr)
This should be a static routine.
> +{
> + int i;
> + struct {
> + u8 version[4];
> + u32 base_addr;
> + } match[] = {
> + {{0x05, 0x00, 0x00, 0x00}, 0x00005000},
> + {{0x05, 0x00, 0x00, 0x01}, 0x00003000},
> + {{0x05, 0x00, 0x00, 0x02}, 0x00003000},
> + {{0x05, 0x00, 0x00, 0x03}, 0x00003000},
> + {{0x05, 0x00, 0x00, 0x05}, 0x00003000}
> + };
> +
> + for (i = 0; i < ARRAY_SIZE(match); ++i)
> + if (bootinfo->hw_version[0] == match[i].version[0] &&
> + bootinfo->hw_version[1] == match[i].version[1] &&
> + bootinfo->hw_version[3] == match[i].version[3]) {
> + *base_addr = match[i].base_addr;
> + return 0;
> + }
> +
> + return -EINVAL;
> +}
> +
> +bool s3fwrn5_fw_is_custom(struct s3fwrn5_fw_cmd_get_bootinfo_rsp *bootinfo)
> +{
> + return !!bootinfo->hw_version[2];
> +}
This should be a static inline routine.
> +static int s3fwrn5_i2c_nci_read(struct s3fwrn5_i2c_phy *phy)
> +{
> + struct nci_ctrl_hdr hdr;
> + struct sk_buff *skb;
> + int ret;
> +
> + ENTER();
> +
> + ret = i2c_master_recv(phy->i2c_dev, (char *)&hdr, NCI_CTRL_HDR_SIZE);
> + if (ret < 0)
> + return ret;
> +
> + if (ret < NCI_CTRL_HDR_SIZE)
> + return -EBADMSG;
> +
> + skb = alloc_skb(NCI_CTRL_HDR_SIZE + hdr.plen, GFP_KERNEL);
> + if (!skb)
> + return -ENOMEM;
> +
> + memcpy(skb_put(skb, NCI_CTRL_HDR_SIZE), &hdr, NCI_CTRL_HDR_SIZE);
> +
> + if (hdr.plen == 0)
> + goto out;
> +
> + ret = i2c_master_recv(phy->i2c_dev, skb_put(skb, hdr.plen), hdr.plen);
> + if (ret != hdr.plen) {
> + kfree_skb(skb);
> + return -EBADMSG;
> + }
> +
> +out:
> + DUMP_SKB("R:", skb);
> +
> + return s3fwrn5_recv_frame(phy->ndev, skb, S3FWRN5_MODE_NCI);
> +}
Please factorize s3fwrn5_i2c_nci_read and s3fwrn5_i2c_fw_read. They
share the same logic, only the header sizes differ.
> +static int s3fwrn5_nci_core_reset(struct s3fwrn5_nci_info *nci_info,
> + struct nci_core_reset_cmd *cmd)
> +{
> + struct nci_core_reset_rsp *nci_rsp;
> + struct sk_buff *msg, *rsp = NULL;
> + int ret;
> +
> + ret = s3fwrn5_nci_prep_cmd(&msg, NCI_OP_CORE_RESET_CMD,
> + sizeof(*cmd), cmd);
> + if (ret < 0)
> + return ret;
> +
> + ret = s3fwrn5_nci_send_msg(nci_info, msg, &rsp);
> + kfree_skb(msg);
> + if (ret < 0)
> + goto out;
> +
> + nci_rsp = (void *) rsp->data + NCI_CTRL_HDR_SIZE;
> + if (nci_rsp->status != NCI_STATUS_OK)
> + return -EIO;
> +
> +out:
> + kfree_skb(rsp);
> + return ret;
> +}
Please export an nci_core_reset() command from nci/core.c for that, that
would be a __nci_request(ndev, nci_reset_req) wrapper.
Your prep_cmd and send_msg routines are a duplication of the nci/core.c
code, we should avoid that.
> +static int s3fwrn5_nci_core_init_get_version(struct s3fwrn5_nci_info *nci_info,
> + __u32 *version)
> +{
> + struct nci_core_init_rsp_1 *nci_rsp;
> + struct sk_buff *msg, *rsp = NULL;
> + struct nci_ctrl_hdr *hdr;
> + int ret;
> +
> + ret = s3fwrn5_nci_prep_cmd(&msg, NCI_OP_CORE_INIT_CMD, 0, NULL);
> + if (ret < 0)
> + return ret;
> +
> + ret = s3fwrn5_nci_send_msg(nci_info, msg, &rsp);
> + kfree_skb(msg);
> + if (ret < 0)
> + goto out;
> +
> + hdr = (void *) rsp->data;
> + if (hdr->plen != 20) {
> + ret = -EIO;
> + goto out;
> + }
> +
> + nci_rsp = (void *) rsp->data + NCI_CTRL_HDR_SIZE;
> + if (nci_rsp->status != NCI_STATUS_OK) {
> + ret = -EIO;
> + goto out;
> + }
> +
> + memcpy(version, rsp->data + NCI_CTRL_HDR_SIZE + 16, 4);
This should all be contained in ndev->manufact_id and
ndev->manufact_specific_info after nci_open_device() is done with the
init path.
If you need a hook (post_setup) for this, please add it and call it from
nci_open_device().
> +static int s3fwrn5_nci_fw_cfg(struct s3fwrn5_nci_info *nci_info,
> + struct nci_prop_fw_cfg_cmd *cmd)
> +{
> + struct nci_prop_fw_cfg_rsp *nci_rsp;
> + struct sk_buff *msg, *rsp = NULL;
> + int ret;
> +
> + ret = s3fwrn5_nci_prep_cmd(&msg, NCI_OP_PROP_FW_CFG_CMD,
> + sizeof(*cmd), cmd);
> + if (ret < 0)
> + return ret;
> +
> + ret = s3fwrn5_nci_send_msg(nci_info, msg, &rsp);
> + kfree_skb(msg);
> + if (ret < 0)
> + goto out;
> +
> + nci_rsp = (void *) rsp->data + NCI_CTRL_HDR_SIZE;
> + if (nci_rsp->status != NCI_STATUS_OK)
> + ret = -EIO;
> +
> +out:
> + kfree_skb(rsp);
> + return ret;
> +}
Please use nci_prop_cmd for sending proprietary and register your
proprietary responses through nci_prop_ops.
See b6355e972aaab0173ce11a1650e7dba67f820918 for more details.
> +int s3fwrn5_nci_recv_frame(struct nci_dev *ndev, struct sk_buff *skb)
> +{
> + struct s3fwrn5_info *info = nci_get_drvdata(ndev);
> + struct s3fwrn5_nci_info *nci_info = &info->nci_info;
> +
> + ENTER();
> +
> + BUG_ON(nci_info->rsp);
> +
> + nci_info->rsp = skb;
> +
> + complete(&nci_info->completion);
> +
> + return 0;
> +}
This one should be entirely handled by the NCI core code.
Cheers,
Samuel.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists