On Wed, Jul 13, 2011 at 12:22:57PM +0300, Dan Carpenter wrote: > Nice. :) > > There are a some smatch warnings you might want to take a look at: > > drivers/net/wireless/ath/ath6kl/sdio.c +373 ath6kl_sdio_setup_scat_resource(42) warn: 's_req' was already freed. > drivers/net/wireless/ath/ath6kl/htc.c +2480 htc_create(74) warn: possible memory leak of 'packet' > drivers/net/wireless/ath/ath6kl/txrx.c +156 ath6kl_powersave_ap(76) error: we previously assumed 'conn' could be null. > (The line number is off on that one. Should be +142. Odd) > drivers/net/wireless/ath/ath6kl/txrx.c +758 ath6kl_alloc_amsdu_rxbuf(33) warn: can 'packet' even be NULL? > drivers/net/wireless/ath/ath5k/phy.c +847 ath5k_hw_rfregs_init(152) error: potential null derefence 'go'. > drivers/net/wireless/ath/ath9k/ar9003_eeprom.c +3320 ar9300_eeprom_restore_internal(20) warn: returning -1 instead of -ENOMEM is sloppy > drivers/net/wireless/ath/ath9k/hif_usb.c +135 hif_usb_mgmt_cb(6) warn: variable dereferenced before check 'cmd' > drivers/net/wireless/ath/ath9k/pci.c +68 ath_pci_eeprom_read(12) error: buffer overflow 'pdata->eeprom_data' 2048 <= 2048 Ah... Half of those are for other atheros drivers... I didn't read the warnings carefully. Btw in drivers/net/wireless/ath/ath9k/debug.c read_file_stations() the we use kmalloc() to allocate a 64000 char buffer. That's a lot of contiguous memory to try allocate. (But obviously that's ath9k and not this drivers)... regards, dan carpenter _______________________________________________ devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxx http://driverdev.linuxdriverproject.org/mailman/listinfo/devel
