forked from Minki/linux
ath10k: fix possible memory leak in new FW loading
Some failpaths did `return` instead of a `goto` leaking requested firmware. Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
This commit is contained in:
parent
fd9c4864a7
commit
9bab1cc010
@ -404,12 +404,14 @@ static int ath10k_core_fetch_firmware_api_n(struct ath10k *ar, const char *name)
|
||||
if (len < magic_len) {
|
||||
ath10k_err("firmware image too small to contain magic: %zu\n",
|
||||
len);
|
||||
return -EINVAL;
|
||||
ret = -EINVAL;
|
||||
goto err;
|
||||
}
|
||||
|
||||
if (memcmp(data, ATH10K_FIRMWARE_MAGIC, magic_len) != 0) {
|
||||
ath10k_err("Invalid firmware magic\n");
|
||||
return -EINVAL;
|
||||
ret = -EINVAL;
|
||||
goto err;
|
||||
}
|
||||
|
||||
/* jump over the padding */
|
||||
@ -431,7 +433,8 @@ static int ath10k_core_fetch_firmware_api_n(struct ath10k *ar, const char *name)
|
||||
if (len < ie_len) {
|
||||
ath10k_err("Invalid length for FW IE %d (%zu < %zu)\n",
|
||||
ie_id, len, ie_len);
|
||||
return -EINVAL;
|
||||
ret = -EINVAL;
|
||||
goto err;
|
||||
}
|
||||
|
||||
switch (ie_id) {
|
||||
|
Loading…
Reference in New Issue
Block a user