Skip to contentSkip to conversation

wilc1000: move wilc driver out of staging

bug report Dan Carpenter 2 participants 2 messages Aug 29, 2025

lore
Share & tools
lore
#0 lore raw

Hello Ajay Singh,

[ Obviously this bug was in staging as well... ]

Commit 5625f965d764 ("wilc1000: move wilc driver out of staging") from Jun 25, 2020 (linux-next), leads to the following Smatch static checker warning:

	drivers/net/wireless/microchip/wilc1000/wlan_cfg.c:184 wilc_wlan_parse_response_frame()
	error: '__memcpy()' 'cfg->s[i]->str' copy overflow (512 vs 65537)

drivers/net/wireless/microchip/wilc1000/wlan_cfg.c

    138 static void wilc_wlan_parse_response_frame(struct wilc *wl, u8 *info, int size)
    139 {
    140         u16 wid;
    141         u32 len = 0, i = 0;
    142         struct wilc_cfg *cfg = &wl->cfg;
    143 
    144         while (size > 0) {
    145                 i = 0;
    146                 wid = get_unaligned_le16(info);
    147 
    148                 switch (FIELD_GET(WILC_WID_TYPE, wid)) {
    149                 case WID_CHAR:
    150                         while (cfg->b[i].id != WID_NIL && cfg->b[i].id != wid)
    151                                 i++;
    152 

This is the rx path and info comes from skb->data so it feels like we need more bounds checking.

	if (info < 5)
		return;

    153                         if (cfg->b[i].id == wid)
    154                                 cfg->b[i].val = info[4];
    155 
    156                         len = 3;
    157                         break;
    158 
    159                 case WID_SHORT:

	if (info < 6)
		return;

    160                         while (cfg->hw[i].id != WID_NIL && cfg->hw[i].id != wid)
    161                                 i++;
    162 
    163                         if (cfg->hw[i].id == wid)
    164                                 cfg->hw[i].val = get_unaligned_le16(&info[4]);
    165 
    166                         len = 4;
    167                         break;
    168 
    169                 case WID_INT:

	if (info < 8)
		return;

    170                         while (cfg->w[i].id != WID_NIL && cfg->w[i].id != wid)
    171                                 i++;
    172 
    173                         if (cfg->w[i].id == wid)
    174                                 cfg->w[i].val = get_unaligned_le32(&info[4]);
    175 
    176                         len = 6;
    177                         break;
    178 
    179                 case WID_STR:

How many bytes are there in cfg->s[i].str? Smatch thinks 512, but I don't know where Smatch is getting that...

	len = 2 + get_unaligned_le16(&info[2]);
	if (len + 2 > SOME_LIMIT)
		return;

    180                         while (cfg->s[i].id != WID_NIL && cfg->s[i].id != wid)
    181                                 i++;
    182 
    183                         if (cfg->s[i].id == wid)
--> 184                                 memcpy(cfg->s[i].str, &info[2],
    185                                        get_unaligned_le16(&info[2]) + 2);
    186 
    187                         len = 2 + get_unaligned_le16(&info[2]);
    188                         break;
    189 
    190                 default:
    191                         break;
    192                 }
    193                 size -= (2 + len);
    194                 info += (2 + len);
    195         }
    196 }

regards,
dan carpenter

#1 lore raw

Hi Dan,

Thanks for pointing out this warning along with the analysis.

Quoted 30 lines · Dan Carpenter

EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe

Hello Ajay Singh,

[ Obviously this bug was in staging as well... ]

Commit 5625f965d764 ("wilc1000: move wilc driver out of staging") from Jun 25, 2020 (linux-next), leads to the following Smatch static checker warning:

        drivers/net/wireless/microchip/wilc1000/wlan_cfg.c:184 wilc_wlan_parse_response_frame()
        error: '__memcpy()' 'cfg->s[i]->str' copy overflow (512 vs 65537)

drivers/net/wireless/microchip/wilc1000/wlan_cfg.c

    138 static void wilc_wlan_parse_response_frame(struct wilc *wl, u8 *info, int size)
    139 {
    140         u16 wid;
    141         u32 len = 0, i = 0;
    142         struct wilc_cfg *cfg = &wl->cfg;
    143
    144         while (size > 0) {
    145                 i = 0;
    146                 wid = get_unaligned_le16(info);
    147
    148                 switch (FIELD_GET(WILC_WID_TYPE, wid)) {
    149                 case WID_CHAR:
    150                         while (cfg->b[i].id != WID_NIL && cfg->b[i].id != wid)
    151                                 i++;
    152
Quoted 6 lines

This is the rx path and info comes from skb->data so it feels like we need more bounds checking.

        if (info < 5)
                return;

Agree, the read can go beyond the buffer limit so it's safe to add boundary checks before accessing the 'info' buffer.

Quoted 36 lines
    153                         if (cfg->b[i].id == wid)
    154                                 cfg->b[i].val = info[4];
    155
    156                         len = 3;
    157                         break;
    158
    159                 case WID_SHORT:

        if (info < 6)
                return;

    160                         while (cfg->hw[i].id != WID_NIL && cfg->hw[i].id != wid)
    161                                 i++;
    162
    163                         if (cfg->hw[i].id == wid)
    164                                 cfg->hw[i].val = get_unaligned_le16(&info[4]);
    165
    166                         len = 4;
    167                         break;
    168
    169                 case WID_INT:

        if (info < 8)
                return;

    170                         while (cfg->w[i].id != WID_NIL && cfg->w[i].id != wid)
    171                                 i++;
    172
    173                         if (cfg->w[i].id == wid)
    174                                 cfg->w[i].val = get_unaligned_le32(&info[4]);
    175
    176                         len = 6;
    177                         break;
    178
    179                 case WID_STR:
Quoted 6 lines

How many bytes are there in cfg->s[i].str? Smatch thinks 512, but I don't know where Smatch is getting that...

        len = 2 + get_unaligned_le16(&info[2]);
        if (len + 2 > SOME_LIMIT)
                return;

I think Smatch got 512 value from the below macro.

#define WILC_MAX_ASSOC_RESP_FRAME_SIZE 512

Patch[1] has been submitted to address this warning. I hope this would resolve the warning.

1.
https://lore.kernel.org/linux-wireless/[email protected]/

Regards,
Ajay