| Message ID | 20191018114321.13131-1-labbott@redhat.com |
|---|---|
| State | Awaiting Upstream |
| Delegated to: | David Miller |
| Headers | show |
| Series | [v2] rtlwifi: Fix potential overflow on P2P code | expand |
On Fri, 2019-10-18 at 07:43 -0400, Laura Abbott wrote: > Nicolas Waisman noticed that even though noa_len is checked for > a compatible length it's still possible to overrun the buffers > of p2pinfo since there's no check on the upper bound of noa_num. > Bound noa_num against P2P_MAX_NOA_NUM. > > Reported-by: Nicolas Waisman <nico@semmle.com> > Signed-off-by: Laura Abbott <labbott@redhat.com> Acked-by: Ping-Ke Shih <pkshih@realtek.com> and Please CC to stable Cc: Stable <stable@vger.kernel.org> # 4.4+ --- Hi Kalle, This bug was existing since v3.10, and directory of wireless drivers were moved at v4.4. Do I need send another patch to fix this issue for longterm kernel v3.16.75? Thanks PK > --- > v2: Use P2P_MAX_NOA_NUM instead of erroring out. > --- > drivers/net/wireless/realtek/rtlwifi/ps.c | 6 ++++++ > 1 file changed, 6 insertions(+) > > diff --git a/drivers/net/wireless/realtek/rtlwifi/ps.c > b/drivers/net/wireless/realtek/rtlwifi/ps.c > index 70f04c2f5b17..fff8dda14023 100644 > --- a/drivers/net/wireless/realtek/rtlwifi/ps.c > +++ b/drivers/net/wireless/realtek/rtlwifi/ps.c > @@ -754,6 +754,9 @@ static void rtl_p2p_noa_ie(struct ieee80211_hw *hw, void > *data, > return; > } else { > noa_num = (noa_len - 2) / 13; > + if (noa_num > P2P_MAX_NOA_NUM) > + noa_num = P2P_MAX_NOA_NUM; > + > } > noa_index = ie[3]; > if (rtlpriv->psc.p2p_ps_info.p2p_ps_mode == > @@ -848,6 +851,9 @@ static void rtl_p2p_action_ie(struct ieee80211_hw *hw, > void *data, > return; > } else { > noa_num = (noa_len - 2) / 13; > + if (noa_num > P2P_MAX_NOA_NUM) > + noa_num = P2P_MAX_NOA_NUM; > + > } > noa_index = ie[3]; > if (rtlpriv->psc.p2p_ps_info.p2p_ps_mode ==
Laura Abbott <labbott@redhat.com> writes: > Nicolas Waisman noticed that even though noa_len is checked for > a compatible length it's still possible to overrun the buffers > of p2pinfo since there's no check on the upper bound of noa_num. > Bound noa_num against P2P_MAX_NOA_NUM. > > Reported-by: Nicolas Waisman <nico@semmle.com> > Signed-off-by: Laura Abbott <labbott@redhat.com> > --- > v2: Use P2P_MAX_NOA_NUM instead of erroring out. > --- > drivers/net/wireless/realtek/rtlwifi/ps.c | 6 ++++++ > 1 file changed, 6 insertions(+) > > diff --git a/drivers/net/wireless/realtek/rtlwifi/ps.c b/drivers/net/wireless/realtek/rtlwifi/ps.c > index 70f04c2f5b17..fff8dda14023 100644 > --- a/drivers/net/wireless/realtek/rtlwifi/ps.c > +++ b/drivers/net/wireless/realtek/rtlwifi/ps.c > @@ -754,6 +754,9 @@ static void rtl_p2p_noa_ie(struct ieee80211_hw *hw, void *data, > return; > } else { > noa_num = (noa_len - 2) / 13; > + if (noa_num > P2P_MAX_NOA_NUM) > + noa_num = P2P_MAX_NOA_NUM; > + > } > noa_index = ie[3]; > if (rtlpriv->psc.p2p_ps_info.p2p_ps_mode == > @@ -848,6 +851,9 @@ static void rtl_p2p_action_ie(struct ieee80211_hw *hw, void *data, > return; > } else { > noa_num = (noa_len - 2) / 13; > + if (noa_num > P2P_MAX_NOA_NUM) > + noa_num = P2P_MAX_NOA_NUM; IMHO using min() would be cleaner, but I'm fine with this as well. Up to you.
Pkshih <pkshih@realtek.com> writes: > On Fri, 2019-10-18 at 07:43 -0400, Laura Abbott wrote: >> Nicolas Waisman noticed that even though noa_len is checked for >> a compatible length it's still possible to overrun the buffers >> of p2pinfo since there's no check on the upper bound of noa_num. >> Bound noa_num against P2P_MAX_NOA_NUM. >> >> Reported-by: Nicolas Waisman <nico@semmle.com> >> Signed-off-by: Laura Abbott <labbott@redhat.com> > > Acked-by: Ping-Ke Shih <pkshih@realtek.com> > and Please CC to stable > Cc: Stable <stable@vger.kernel.org> # 4.4+ > > --- > > Hi Kalle, > > This bug was existing since v3.10, and directory of wireless drivers were > moved at v4.4. Do I need send another patch to fix this issue for longterm > kernel v3.16.75? Yeah, I think you need to send a separate patch to the stable list (after this commit is in Linus' tree).
On 10/19/19 6:51 AM, Kalle Valo wrote: > Laura Abbott <labbott@redhat.com> writes: > >> Nicolas Waisman noticed that even though noa_len is checked for >> a compatible length it's still possible to overrun the buffers >> of p2pinfo since there's no check on the upper bound of noa_num. >> Bound noa_num against P2P_MAX_NOA_NUM. >> >> Reported-by: Nicolas Waisman <nico@semmle.com> >> Signed-off-by: Laura Abbott <labbott@redhat.com> >> --- >> v2: Use P2P_MAX_NOA_NUM instead of erroring out. >> --- >> drivers/net/wireless/realtek/rtlwifi/ps.c | 6 ++++++ >> 1 file changed, 6 insertions(+) >> >> diff --git a/drivers/net/wireless/realtek/rtlwifi/ps.c b/drivers/net/wireless/realtek/rtlwifi/ps.c >> index 70f04c2f5b17..fff8dda14023 100644 >> --- a/drivers/net/wireless/realtek/rtlwifi/ps.c >> +++ b/drivers/net/wireless/realtek/rtlwifi/ps.c >> @@ -754,6 +754,9 @@ static void rtl_p2p_noa_ie(struct ieee80211_hw *hw, void *data, >> return; >> } else { >> noa_num = (noa_len - 2) / 13; >> + if (noa_num > P2P_MAX_NOA_NUM) >> + noa_num = P2P_MAX_NOA_NUM; >> + >> } >> noa_index = ie[3]; >> if (rtlpriv->psc.p2p_ps_info.p2p_ps_mode == >> @@ -848,6 +851,9 @@ static void rtl_p2p_action_ie(struct ieee80211_hw *hw, void *data, >> return; >> } else { >> noa_num = (noa_len - 2) / 13; >> + if (noa_num > P2P_MAX_NOA_NUM) >> + noa_num = P2P_MAX_NOA_NUM; > > IMHO using min() would be cleaner, but I'm fine with this as well. Up to > you. > I believe the intention is to re-write this anyway so I'd prefer to just get this in given the uptick this issue seems to have gotten. Thanks, Laura
Laura Abbott <labbott@redhat.com> writes: > On 10/19/19 6:51 AM, Kalle Valo wrote: >> Laura Abbott <labbott@redhat.com> writes: >> >>> Nicolas Waisman noticed that even though noa_len is checked for >>> a compatible length it's still possible to overrun the buffers >>> of p2pinfo since there's no check on the upper bound of noa_num. >>> Bound noa_num against P2P_MAX_NOA_NUM. >>> >>> Reported-by: Nicolas Waisman <nico@semmle.com> >>> Signed-off-by: Laura Abbott <labbott@redhat.com> >>> --- >>> v2: Use P2P_MAX_NOA_NUM instead of erroring out. >>> --- >>> drivers/net/wireless/realtek/rtlwifi/ps.c | 6 ++++++ >>> 1 file changed, 6 insertions(+) >>> >>> diff --git a/drivers/net/wireless/realtek/rtlwifi/ps.c b/drivers/net/wireless/realtek/rtlwifi/ps.c >>> index 70f04c2f5b17..fff8dda14023 100644 >>> --- a/drivers/net/wireless/realtek/rtlwifi/ps.c >>> +++ b/drivers/net/wireless/realtek/rtlwifi/ps.c >>> @@ -754,6 +754,9 @@ static void rtl_p2p_noa_ie(struct ieee80211_hw *hw, void *data, >>> return; >>> } else { >>> noa_num = (noa_len - 2) / 13; >>> + if (noa_num > P2P_MAX_NOA_NUM) >>> + noa_num = P2P_MAX_NOA_NUM; >>> + >>> } >>> noa_index = ie[3]; >>> if (rtlpriv->psc.p2p_ps_info.p2p_ps_mode == >>> @@ -848,6 +851,9 @@ static void rtl_p2p_action_ie(struct ieee80211_hw *hw, void *data, >>> return; >>> } else { >>> noa_num = (noa_len - 2) / 13; >>> + if (noa_num > P2P_MAX_NOA_NUM) >>> + noa_num = P2P_MAX_NOA_NUM; >> >> IMHO using min() would be cleaner, but I'm fine with this as well. Up to >> you. >> > > I believe the intention is to re-write this anyway so I'd prefer to > just get this in given the uptick this issue seems to have gotten. Ok, I'll queue this to v5.4.
Laura Abbott <labbott@redhat.com> wrote: > Nicolas Waisman noticed that even though noa_len is checked for > a compatible length it's still possible to overrun the buffers > of p2pinfo since there's no check on the upper bound of noa_num. > Bound noa_num against P2P_MAX_NOA_NUM. > > Reported-by: Nicolas Waisman <nico@semmle.com> > Signed-off-by: Laura Abbott <labbott@redhat.com> > Acked-by: Ping-Ke Shih <pkshih@realtek.com> Patch applied to wireless-drivers.git, thanks. 8c55dedb795b rtlwifi: Fix potential overflow on P2P code
diff --git a/drivers/net/wireless/realtek/rtlwifi/ps.c b/drivers/net/wireless/realtek/rtlwifi/ps.c index 70f04c2f5b17..fff8dda14023 100644 --- a/drivers/net/wireless/realtek/rtlwifi/ps.c +++ b/drivers/net/wireless/realtek/rtlwifi/ps.c @@ -754,6 +754,9 @@ static void rtl_p2p_noa_ie(struct ieee80211_hw *hw, void *data, return; } else { noa_num = (noa_len - 2) / 13; + if (noa_num > P2P_MAX_NOA_NUM) + noa_num = P2P_MAX_NOA_NUM; + } noa_index = ie[3]; if (rtlpriv->psc.p2p_ps_info.p2p_ps_mode == @@ -848,6 +851,9 @@ static void rtl_p2p_action_ie(struct ieee80211_hw *hw, void *data, return; } else { noa_num = (noa_len - 2) / 13; + if (noa_num > P2P_MAX_NOA_NUM) + noa_num = P2P_MAX_NOA_NUM; + } noa_index = ie[3]; if (rtlpriv->psc.p2p_ps_info.p2p_ps_mode ==
Nicolas Waisman noticed that even though noa_len is checked for a compatible length it's still possible to overrun the buffers of p2pinfo since there's no check on the upper bound of noa_num. Bound noa_num against P2P_MAX_NOA_NUM. Reported-by: Nicolas Waisman <nico@semmle.com> Signed-off-by: Laura Abbott <labbott@redhat.com> --- v2: Use P2P_MAX_NOA_NUM instead of erroring out. --- drivers/net/wireless/realtek/rtlwifi/ps.c | 6 ++++++ 1 file changed, 6 insertions(+)