• Ping-Ke Shih's avatar
    rtw88: Fix array overrun in rtw_get_tx_power_params() · 2ff25985
    Ping-Ke Shih authored
    Using a kernel with the Undefined Behaviour Sanity Checker (UBSAN) enabled, the
    following array overrun is logged:
    
    ================================================================================
    UBSAN: array-index-out-of-bounds in /home/finger/wireless-drivers-next/drivers/net/wireless/realtek/rtw88/phy.c:1789:34
    index 5 is out of range for type 'u8 [5]'
    CPU: 2 PID: 84 Comm: kworker/u16:3 Tainted: G           O      5.12.0-rc5-00086-gd88bba47038e-dirty #651
    Hardware name: TOSHIBA TECRA A50-A/TECRA A50-A, BIOS Version 4.50   09/29/2014
    Workqueue: phy0 ieee80211_scan_work [mac80211]
    Call Trace:
     dump_stack+0x64/0x7c
     ubsan_epilogue+0x5/0x40
     __ubsan_handle_out_of_bounds.cold+0x43/0x48
     rtw_get_tx_power_params+0x83a/drivers/net/wireless/realtek/rtw88/0xad0 [rtw_core]
     ? rtw_pci_read16+0x20/0x20 [rtw_pci]
     ? check_hw_ready+0x50/0x90 [rtw_core]
     rtw_phy_get_tx_power_index+0x4d/0xd0 [rtw_core]
     rtw_phy_set_tx_power_level+0xee/0x1b0 [rtw_core]
     rtw_set_channel+0xab/0x110 [rtw_core]
     rtw_ops_config+0x87/0xc0 [rtw_core]
     ieee80211_hw_config+0x9d/0x130 [mac80211]
     ieee80211_scan_state_set_channel+0x81/0x170 [mac80211]
     ieee80211_scan_work+0x19f/0x2a0 [mac80211]
     process_one_work+0x1dd/0x3a0
     worker_thread+0x49/0x330
     ? rescuer_thread+0x3a0/0x3a0
     kthread+0x134/0x150
     ? kthread_create_worker_on_cpu+0x70/0x70
     ret_from_fork+0x22/0x30
    ================================================================================
    
    The statement where an array is being overrun is shown in the following snippet:
    
    	if (rate <= DESC_RATE11M)
    		tx_power = pwr_idx_2g->cck_base[group];
    	else
    ====>		tx_power = pwr_idx_2g->bw40_base[group];
    
    The associated arrays are defined in main.h as follows:
    
    struct rtw_2g_txpwr_idx {
    	u8 cck_base[6];
    	u8 bw40_base[5];
    	struct rtw_2g_1s_pwr_idx_diff ht_1s_diff;
    	struct rtw_2g_ns_pwr_idx_diff ht_2s_diff;
    	struct rtw_2g_ns_pwr_idx_diff ht_3s_diff;
    	struct rtw_2g_ns_pwr_idx_diff ht_4s_diff;
    };
    
    The problem arises because the value of group is 5 for channel 14. The trivial
    increase in the dimension of bw40_base fails as this struct must match the layout of
    efuse. The fix is to add the rate as an argument to rtw_get_channel_group() and set
    the group for channel 14 to 4 if rate <= DESC_RATE11M.
    
    This patch fixes commit fa6dfe6b ("rtw88: resolve order of tx power setting routines")
    
    Fixes: fa6dfe6b
    
     ("rtw88: resolve order of tx power setting routines")
    Reported-by: default avatarБогдан Пилипенко <bogdan.pylypenko107@gmail.com>
    Signed-off-by: default avatarLarry Finger <Larry.Finger@lwfinger.net>
    Signed-off-by: default avatarPing-Ke Shih <pkshih@realtek.com>
    Cc: Stable <stable@vger.kernel.org>
    Signed-off-by: default avatarKalle Valo <kvalo@codeaurora.org>
    Link: https://lore.kernel.org/r/20210401192717.28927-1-Larry.Finger@lwfinger.net
    2ff25985
phy.c 55.7 KB