mbox series

[v3,00/18] wilc1000: move out of staging

Message ID 20200225074105.7740-1-ajay.kathat@microchip.com (mailing list archive)
Headers show
Series wilc1000: move out of staging | expand

Message

Ajay Singh Feb. 25, 2020, 7:41 a.m. UTC
From: Ajay Singh <ajay.kathat@microchip.com>

This patch series is to review and move wilc1000 driver out of staging.
Most of the review comments received in [1] & [2] are addressed in the
latest code. Please review and provide your inputs.

[1]. https://lore.kernel.org/linux-wireless/1537957525-11467-1-git-send-email-ajay.kathat@microchip.com/
[2]. https://lore.kernel.org/linux-wireless/1562896697-8002-1-git-send-email-ajay.kathat@microchip.com/

Changes since v2:
  - use 'struct' to extract FW info from received commands.
  - make use of C style comments instead of C++.
  - remove use of bool type for firmware struct.
  - deleted unused code related to interrupt handling.
  - make use of RCU list to maintain interfaces list.
  - remove 'wilc_' prefix from file name.
  - added 'WILC_' prefix for header guard macro.
  - remove use of infinite loops(i.e. while(1)).
  - move firmware realted struct to a separate file.
  - refactor SPI command handling by using 'struct'.
  - use different functions to handle different SPI commands.
  - remove use of vendor specific IE for p2p handling.
  - refactor p2p related code to avoid use of buf pointer operation.
  - make use of FIELD_GET/PREP macro.
  - use #define instead of magic values.
  - use YAML schemes for DT binding documentation.
  - deleted unused code from spi.c and sdio.c.
  - added changes for few issues reported by smatch static code analyzer.

Changes since v1:
  - remove use of shadow buffer to keep scan result.
  - remove internal messaging flow to handle cfg80211_ops.
  - make use of cfg80211 provide API.
  - use 'struct' for packing firmware commands.
  - make use of kernel API's and Macro.
  - remove unnecessary log messages
  - supported dynamically add/remove interfaces.
  - cleanup and deleted around 3.3k lines of code.


Ajay Singh (18):
  wilc1000: add hif.h
  wilc1000: add hif.c
  wilc1000: add wlan_if.h
  wilc1000: add wlan_cfg.h
  wilc1000: add wlan_cfg.c
  wilc1000: add cfg80211.c
  wilc1000: add cfg80211.h
  wilc1000: add netdev.h
  wilc1000: add netdev.c
  wilc1000: add mon.c
  wilc1000: add spi.c
  wilc1000: add wlan.h
  wilc1000: add wlan.c
  wilc1000: add sdio.c
  wilc1000: add fw.h
  dt: bindings: net: add microchip,wilc1000,sdio.yaml
  dt: bindings: net: add microchip,wilc1000,spi.yaml
  wilc1000: add Makefile and Kconfig files for wilc1000 compilation

 .../net/wireless/microchip,wilc1000,sdio.yaml |   68 +
 .../net/wireless/microchip,wilc1000,spi.yaml  |   61 +
 drivers/net/wireless/Kconfig                  |    1 +
 drivers/net/wireless/Makefile                 |    1 +
 drivers/net/wireless/microchip/Kconfig        |   15 +
 drivers/net/wireless/microchip/Makefile       |    2 +
 .../net/wireless/microchip/wilc1000/Kconfig   |   42 +
 .../net/wireless/microchip/wilc1000/Makefile  |   14 +
 .../wireless/microchip/wilc1000/cfg80211.c    | 1852 ++++++++++++++++
 .../wireless/microchip/wilc1000/cfg80211.h    |   29 +
 drivers/net/wireless/microchip/wilc1000/fw.h  |  119 +
 drivers/net/wireless/microchip/wilc1000/hif.c | 1959 +++++++++++++++++
 drivers/net/wireless/microchip/wilc1000/hif.h |  214 ++
 drivers/net/wireless/microchip/wilc1000/mon.c |  260 +++
 .../net/wireless/microchip/wilc1000/netdev.c  |  940 ++++++++
 .../net/wireless/microchip/wilc1000/netdev.h  |  295 +++
 .../net/wireless/microchip/wilc1000/sdio.c    | 1030 +++++++++
 drivers/net/wireless/microchip/wilc1000/spi.c | 1001 +++++++++
 .../net/wireless/microchip/wilc1000/wlan.c    | 1240 +++++++++++
 .../net/wireless/microchip/wilc1000/wlan.h    |  398 ++++
 .../wireless/microchip/wilc1000/wlan_cfg.c    |  413 ++++
 .../wireless/microchip/wilc1000/wlan_cfg.h    |   54 +
 .../net/wireless/microchip/wilc1000/wlan_if.h |  803 +++++++
 drivers/staging/Kconfig                       |    2 -
 drivers/staging/Makefile                      |    1 -
 25 files changed, 10811 insertions(+), 3 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/net/wireless/microchip,wilc1000,sdio.yaml
 create mode 100644 Documentation/devicetree/bindings/net/wireless/microchip,wilc1000,spi.yaml
 create mode 100644 drivers/net/wireless/microchip/Kconfig
 create mode 100644 drivers/net/wireless/microchip/Makefile
 create mode 100644 drivers/net/wireless/microchip/wilc1000/Kconfig
 create mode 100644 drivers/net/wireless/microchip/wilc1000/Makefile
 create mode 100644 drivers/net/wireless/microchip/wilc1000/cfg80211.c
 create mode 100644 drivers/net/wireless/microchip/wilc1000/cfg80211.h
 create mode 100644 drivers/net/wireless/microchip/wilc1000/fw.h
 create mode 100644 drivers/net/wireless/microchip/wilc1000/hif.c
 create mode 100644 drivers/net/wireless/microchip/wilc1000/hif.h
 create mode 100644 drivers/net/wireless/microchip/wilc1000/mon.c
 create mode 100644 drivers/net/wireless/microchip/wilc1000/netdev.c
 create mode 100644 drivers/net/wireless/microchip/wilc1000/netdev.h
 create mode 100644 drivers/net/wireless/microchip/wilc1000/sdio.c
 create mode 100644 drivers/net/wireless/microchip/wilc1000/spi.c
 create mode 100644 drivers/net/wireless/microchip/wilc1000/wlan.c
 create mode 100644 drivers/net/wireless/microchip/wilc1000/wlan.h
 create mode 100644 drivers/net/wireless/microchip/wilc1000/wlan_cfg.c
 create mode 100644 drivers/net/wireless/microchip/wilc1000/wlan_cfg.h
 create mode 100644 drivers/net/wireless/microchip/wilc1000/wlan_if.h

Comments

Dan Carpenter March 2, 2020, 9:23 a.m. UTC | #1
There are a few static checker warnings from Friday's linux-next.  Only
the first one is important.  (Not all these Smatch warnings have been
published).

drivers/staging/wilc1000/hif.c:804 wilc_hif_pack_sta_param() warn: '&params->ht_capa' sometimes too small '8' size = 29

drivers/staging/wilc1000/hif.c
   787  static void wilc_hif_pack_sta_param(u8 *cur_byte, const u8 *mac,
   788                                      struct station_parameters *params)
   789  {
   790          ether_addr_copy(cur_byte, mac);
   791          cur_byte += ETH_ALEN;
   792  
   793          put_unaligned_le16(params->aid, cur_byte);
   794          cur_byte += 2;
   795  
   796          *cur_byte++ = params->supported_rates_len;
   797          if (params->supported_rates_len > 0)
   798                  memcpy(cur_byte, params->supported_rates,
   799                         params->supported_rates_len);
   800          cur_byte += params->supported_rates_len;
   801  
   802          if (params->ht_capa) {
   803                  *cur_byte++ = true;
   804                  memcpy(cur_byte, &params->ht_capa,
                                         ^^^^^^^^^^^^^^^^
This is copying the wrong data.  The "&" is wrong.

   805                         sizeof(struct ieee80211_ht_cap));
   806          } else {
   807                  *cur_byte++ = false;
   808          }
   809          cur_byte += sizeof(struct ieee80211_ht_cap);
   810  
   811          put_unaligned_le16(params->sta_flags_mask, cur_byte);
   812          cur_byte += 2;
   813          put_unaligned_le16(params->sta_flags_set, cur_byte);
   814  }


drivers/staging/wilc1000/cfg80211.c:904 del_pmksa() warn: 'i < priv->pmkid_list.numpmkid' 'true' implies 'priv->pmkid_list.numpmkid > 0' is 'true'

drivers/staging/wilc1000/cfg80211.c
   887  static int del_pmksa(struct wiphy *wiphy, struct net_device *netdev,
   888                       struct cfg80211_pmksa *pmksa)
   889  {
   890          u32 i;
   891          int ret = 0;
   892          struct wilc_vif *vif = netdev_priv(netdev);
   893          struct wilc_priv *priv = &vif->priv;
   894  
   895          for (i = 0; i < priv->pmkid_list.numpmkid; i++) {
   896                  if (!memcmp(pmksa->bssid, priv->pmkid_list.pmkidlist[i].bssid,
   897                              ETH_ALEN)) {
   898                          memset(&priv->pmkid_list.pmkidlist[i], 0,
   899                                 sizeof(struct wilc_pmkid));
   900                          break;
   901                  }
   902          }
   903  
   904          if (i < priv->pmkid_list.numpmkid && priv->pmkid_list.numpmkid > 0) {
                                                     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
This part of the condition is a given (must be true).  Delete it.  It's
better to reverse the test and say:

	if (i == priv->pmkid_list.numpmkid)
		return -EINVAL;

   905                  for (; i < (priv->pmkid_list.numpmkid - 1); i++) {
   906                          memcpy(priv->pmkid_list.pmkidlist[i].bssid,
   907                                 priv->pmkid_list.pmkidlist[i + 1].bssid,
   908                                 ETH_ALEN);
   909                          memcpy(priv->pmkid_list.pmkidlist[i].pmkid,
   910                                 priv->pmkid_list.pmkidlist[i + 1].pmkid,
   911                                 WLAN_PMKID_LEN);
   912                  }
   913                  priv->pmkid_list.numpmkid--;
   914          } else {
   915                  ret = -EINVAL;
   916          }
   917  
   918          return ret;
   919  }


drivers/staging/wilc1000/wlan.c:706 wilc_wlan_handle_rx_buff() warn: 'pkt_len' 'true' implies 'pkt_len > 0' is 'true'

drivers/staging/wilc1000/wlan.c
   686          int is_cfg_packet;
   687          u8 *buff_ptr;
   688  
   689          do {
   690                  buff_ptr = buffer + offset;
   691                  header = get_unaligned_le32(buff_ptr);
   692  
   693                  is_cfg_packet = FIELD_GET(WILC_PKT_HDR_CONFIG_FIELD, header);
   694                  pkt_offset = FIELD_GET(WILC_PKT_HDR_OFFSET_FIELD, header);
   695                  tp_len = FIELD_GET(WILC_PKT_HDR_TOTAL_LEN_FIELD, header);
   696                  pkt_len = FIELD_GET(WILC_PKT_HDR_LEN_FIELD, header);
   697  
   698                  if (pkt_len == 0 || tp_len == 0)
                            ^^^^^^^^^^^^

   699                          break;
   700  
   701                  if (pkt_offset & IS_MANAGMEMENT) {
   702                          buff_ptr += HOST_HDR_OFFSET;
   703                          wilc_wfi_mgmt_rx(wilc, buff_ptr, pkt_len);
   704                  } else {
   705                          if (!is_cfg_packet) {
   706                                  if (pkt_len > 0) {
                                            ^^^^^^^^^^^
Delete.

   707                                          wilc_frmw_to_host(wilc, buff_ptr,
   708                                                            pkt_len, pkt_offset);
   709                                  }
   710                          } else {
   711                                  struct wilc_cfg_rsp rsp;
   712  
   713                                  buff_ptr += pkt_offset;
   714  

regards,
dan carpenter
Ajay Singh March 2, 2020, 4:31 p.m. UTC | #2
Hi Dan,


On 02/03/20 2:53 pm, Dan Carpenter wrote:
> EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe
> 
> There are a few static checker warnings from Friday's linux-next.  Only
> the first one is important.  (Not all these Smatch warnings have been
> published).

Thanks. I have submitted a patch series to address Smatch static checker
warnings in staging ml. I will send v4 for this series by including
those changes.

Regards
Ajay