Skip to contentSkip to conversation

staging: Add rtl8723bs sdio wifi driver

bug report Dan Carpenter 2 participants 3 messages Oct 5, 2021 – Oct 11, 2021 Staging

lore
Share & tools
lore
#0 lore raw

Hello Hans de Goede,

The patch 554c0a3abf21: "staging: Add rtl8723bs sdio wifi driver" from Mar 29, 2017, leads to the following Smatch static checker warnings:

drivers/staging/rtl8723bs/core/rtw_security.c:1404 rtw_BIP_verify()
warn: not copying enough bytes for '&le_tmp64' (8 vs 6 bytes)

drivers/staging/rtl8723bs/core/rtw_mlme_ext.c:4058 collect_bss_info()
warn: not copying enough bytes for '&le32_tmp' (4 vs 2 bytes)

drivers/staging/rtl8723bs/core/rtw_security.c

    1372 u32 rtw_BIP_verify(struct adapter *padapter, u8 *precvframe)
    1373 {
    1374         struct rx_pkt_attrib *pattrib = &((union recv_frame *)precvframe)->u.hdr.attrib;
    1375         u8 *pframe;
    1376         u8 *BIP_AAD, *p;
    1377         u32 res = _FAIL;
    1378         uint len, ori_len;
    1379         struct ieee80211_hdr *pwlanhdr;
    1380         u8 mic[16];
    1381         struct mlme_ext_priv *pmlmeext = &padapter->mlmeextpriv;
    1382         __le16 le_tmp;
    1383         __le64 le_tmp64;
                 ^^^^^^^^^^^^^^^

    1384 
    1385         ori_len = pattrib->pkt_len-WLAN_HDR_A3_LEN+BIP_AAD_SIZE;
    1386         BIP_AAD = rtw_zmalloc(ori_len);
    1387 
    1388         if (!BIP_AAD)
    1389                 return _FAIL;
    1390 
    1391         /* PKT start */
    1392         pframe = (unsigned char *)((union recv_frame *)precvframe)->u.hdr.rx_data;
    1393         /* mapping to wlan header */
    1394         pwlanhdr = (struct ieee80211_hdr *)pframe;
    1395         /* save the frame body + MME */
    1396         memcpy(BIP_AAD+BIP_AAD_SIZE, pframe+WLAN_HDR_A3_LEN, pattrib->pkt_len-WLAN_HDR_A3_LEN);
    1397         /* find MME IE pointer */
    1398         p = rtw_get_ie(BIP_AAD+BIP_AAD_SIZE, WLAN_EID_MMIE, &len, pattrib->pkt_len-WLAN_HDR_A3_LEN);
    1399         /* Baron */
    1400         if (p) {
    1401                 u16 keyid = 0;
    1402                 u64 temp_ipn = 0;
    1403                 /* save packet number */
--> 1404                 memcpy(&le_tmp64, p+4, 6);
                                ^^^^^^^^^^^^^^^^^
    1405                 temp_ipn = le64_to_cpu(le_tmp64);
                                                ^^^^^^^^

This code is copying 6 bytes into a u64 and then treating it as little endian data. The problem is that the last two bytes are uninitialized garbage data. I think if we set "__le64 le_tmp64 = 0;" at the top that would probably work, right?

I could have sent a patch but this code is weird enough that I was hoping for a second opinion.

The bug in collect_bss_info() is a similar uninitialized data issue.

    1406                 /* BIP packet number should bigger than previous BIP packet */
    1407                 if (temp_ipn <= pmlmeext->mgnt_80211w_IPN_rx)
    1408                         goto BIP_exit;
    1409 
    1410                 /* copy key index */
    1411                 memcpy(&le_tmp, p+2, 2);

But this part seems totally wrong again because we haven't incremented p. p + 10?

    1412                 keyid = le16_to_cpu(le_tmp);
    1413                 if (keyid != padapter->securitypriv.dot11wBIPKeyid)
    1414                         goto BIP_exit;
    1415 
    1416                 /* clear the MIC field of MME to zero */
    1417                 memset(p+2+len-8, 0, 8);
    1418 
    1419                 /* conscruct AAD, copy frame control field */
    1420                 memcpy(BIP_AAD, &pwlanhdr->frame_control, 2);
    1421                 ClearRetry(BIP_AAD);
    1422                 ClearPwrMgt(BIP_AAD);
    1423                 ClearMData(BIP_AAD);
    1424                 /* conscruct AAD, copy address 1 to address 3 */
    1425                 memcpy(BIP_AAD+2, pwlanhdr->addr1, 18);
    1426 
    1427                 if (omac1_aes_128(padapter->securitypriv.dot11wBIPKey[padapter->securitypriv.dot11wBIPKeyid].skey
    1428                         , BIP_AAD, ori_len, mic))
    1429                         goto BIP_exit;
    1430 
    1431                 /* MIC field should be last 8 bytes of packet (packet without FCS) */
    1432                 if (!memcmp(mic, pframe+pattrib->pkt_len-8, 8)) {
    1433                         pmlmeext->mgnt_80211w_IPN_rx = temp_ipn;
    1434                         res = _SUCCESS;
    1435                 } else {
    1436                 }
    1437 
    1438         } else {
    1439                 res = RTW_RX_HANDLED;
    1440         }
    1441 BIP_exit:
    1442 
    1443         kfree(BIP_AAD);
    1444         return res;
    1445 }

regards,
dan carpenter

#1 lore raw

Hi Dan,

thank you for your report,

Quoted 56 lines · Dan Carpenter

Hello Hans de Goede,

The patch 554c0a3abf21: "staging: Add rtl8723bs sdio wifi driver" from Mar 29, 2017, leads to the following Smatch static checker warnings:

drivers/staging/rtl8723bs/core/rtw_security.c:1404 rtw_BIP_verify()
warn: not copying enough bytes for '&le_tmp64' (8 vs 6 bytes)

drivers/staging/rtl8723bs/core/rtw_mlme_ext.c:4058 collect_bss_info()
warn: not copying enough bytes for '&le32_tmp' (4 vs 2 bytes)

drivers/staging/rtl8723bs/core/rtw_security.c

    1372 u32 rtw_BIP_verify(struct adapter *padapter, u8 *precvframe)
    1373 {
    1374         struct rx_pkt_attrib *pattrib = &((union recv_frame *)precvframe)->u.hdr.attrib;
    1375         u8 *pframe;
    1376         u8 *BIP_AAD, *p;
    1377         u32 res = _FAIL;
    1378         uint len, ori_len;
    1379         struct ieee80211_hdr *pwlanhdr;
    1380         u8 mic[16];
    1381         struct mlme_ext_priv *pmlmeext = &padapter->mlmeextpriv;
    1382         __le16 le_tmp;
    1383         __le64 le_tmp64;
                 ^^^^^^^^^^^^^^^

    1384 
    1385         ori_len = pattrib->pkt_len-WLAN_HDR_A3_LEN+BIP_AAD_SIZE;
    1386         BIP_AAD = rtw_zmalloc(ori_len);
    1387 
    1388         if (!BIP_AAD)
    1389                 return _FAIL;
    1390 
    1391         /* PKT start */
    1392         pframe = (unsigned char *)((union recv_frame *)precvframe)->u.hdr.rx_data;
    1393         /* mapping to wlan header */
    1394         pwlanhdr = (struct ieee80211_hdr *)pframe;
    1395         /* save the frame body + MME */
    1396         memcpy(BIP_AAD+BIP_AAD_SIZE, pframe+WLAN_HDR_A3_LEN, pattrib->pkt_len-WLAN_HDR_A3_LEN);
    1397         /* find MME IE pointer */
    1398         p = rtw_get_ie(BIP_AAD+BIP_AAD_SIZE, WLAN_EID_MMIE, &len, pattrib->pkt_len-WLAN_HDR_A3_LEN);
    1399         /* Baron */
    1400         if (p) {
    1401                 u16 keyid = 0;
    1402                 u64 temp_ipn = 0;
    1403                 /* save packet number */
--> 1404                 memcpy(&le_tmp64, p+4, 6);
                                ^^^^^^^^^^^^^^^^^
    1405                 temp_ipn = le64_to_cpu(le_tmp64);
                                                ^^^^^^^^

This code is copying 6 bytes into a u64 and then treating it as little endian data. The problem is that the last two bytes are uninitialized garbage data. I think if we set "__le64 le_tmp64 = 0;" at the top that would probably work, right?

Quoted 4 lines

I could have sent a patch but this code is weird enough that I was hoping for a second opinion.

The bug in collect_bss_info() is a similar uninitialized data issue.

You are right I think that it's safer to intitalize to zero le_tmp64 and le32_tmp.

Quoted 10 lines
    1406                 /* BIP packet number should bigger than previous BIP packet */
    1407                 if (temp_ipn <= pmlmeext->mgnt_80211w_IPN_rx)
    1408                         goto BIP_exit;
    1409 
    1410                 /* copy key index */
    1411                 memcpy(&le_tmp, p+2, 2);

But this part seems totally wrong again because we haven't incremented p. p + 10?

I don't know what you mean. I guess that you are adressing the code above (lines 1406-1411).

Anyway I think the code it's right. MMIE layout is:

	1 byte element_id;
	1 byte length;
	2 byte key_id;
	6 byte IPN;
	8 byte MIC;

so to access key_id I have to increment p by 2.

Quoted 36 lines
    1412                 keyid = le16_to_cpu(le_tmp);
    1413                 if (keyid != padapter->securitypriv.dot11wBIPKeyid)
    1414                         goto BIP_exit;
    1415 
    1416                 /* clear the MIC field of MME to zero */
    1417                 memset(p+2+len-8, 0, 8);
    1418 
    1419                 /* conscruct AAD, copy frame control field */
    1420                 memcpy(BIP_AAD, &pwlanhdr->frame_control, 2);
    1421                 ClearRetry(BIP_AAD);
    1422                 ClearPwrMgt(BIP_AAD);
    1423                 ClearMData(BIP_AAD);
    1424                 /* conscruct AAD, copy address 1 to address 3 */
    1425                 memcpy(BIP_AAD+2, pwlanhdr->addr1, 18);
    1426 
    1427                 if (omac1_aes_128(padapter->securitypriv.dot11wBIPKey[padapter->securitypriv.dot11wBIPKeyid].skey
    1428                         , BIP_AAD, ori_len, mic))
    1429                         goto BIP_exit;
    1430 
    1431                 /* MIC field should be last 8 bytes of packet (packet without FCS) */
    1432                 if (!memcmp(mic, pframe+pattrib->pkt_len-8, 8)) {
    1433                         pmlmeext->mgnt_80211w_IPN_rx = temp_ipn;
    1434                         res = _SUCCESS;
    1435                 } else {
    1436                 }
    1437 
    1438         } else {
    1439                 res = RTW_RX_HANDLED;
    1440         }
    1441 BIP_exit:
    1442 
    1443         kfree(BIP_AAD);
    1444         return res;
    1445 }
Quoted 3 lines

regards,
dan carpenter

thank you,

fabio

#2 lore raw
Quoted 15 lines · Fabio Aiuto
Quoted 9 lines
    1406                 /* BIP packet number should bigger than previous BIP packet */
    1407                 if (temp_ipn <= pmlmeext->mgnt_80211w_IPN_rx)
    1408                         goto BIP_exit;
    1409 
    1410                 /* copy key index */
    1411                 memcpy(&le_tmp, p+2, 2);

But this part seems totally wrong again because we haven't incremented p. p + 10?

I don't know what you mean. I guess that you are adressing the code above (lines 1406-1411).

Anyway I think the code it's right. MMIE layout is:

Quoted 8 lines
	1 byte element_id;
	1 byte length;
	2 byte key_id;
	6 byte IPN;
	8 byte MIC;

so to access key_id I have to increment p by 2.

Oh, yes. You're right. I think I got confused because we are not parsing the MMIE layout in order. 1, 2, 3 4. It goes 4 then 2.

regards,
dan carpenter