From: Leon Romanovsky leonro@nvidia.com
[ Upstream commit e86212b6b13a20c5ad404c5597933f57fd0f1519 ]
Users can set any seq/seq_hi/oseq/oseq_hi values. The XFRM core code doesn't prevent from them to set even 0xFFFFFFFF, however this value will cause for traffic drop.
Is is happening because SEQ numbers here mean that packet with such number was processed and next number should be sent on the wire. In this case, the next number will be 0, and it means overflow which causes to (expected) packet drops.
While it can be considered as misconfiguration and handled by XFRM datapath in the same manner as any other SEQ number, let's add validation to easy for packet offloads implementations which need to configure HW with next SEQ to send and not with current SEQ like it is done in core code.
Signed-off-by: Leon Romanovsky leonro@nvidia.com Signed-off-by: Steffen Klassert steffen.klassert@secunet.com Signed-off-by: Sasha Levin sashal@kernel.org ---
**YES** This commit should be backported to stable kernel trees. Here's my extensive analysis: ## Bug Analysis The commit addresses a **user input validation bug** in the XFRM (IPsec transformation) subsystem. Specifically: 1. **Root Cause**: Users can set sequence numbers (`seq`, `seq_hi`, `oseq`, `oseq_hi`) to `0xFFFFFFFF` (U32_MAX) via netlink interface 2. **Problem**: When a sequence number is set to U32_MAX, the next packet will have sequence number 0, which triggers sequence number overflow detection and causes packet drops 3. **Impact**: This leads to **immediate traffic disruption** for IPsec connections ## Code Changes Analysis The fix adds **comprehensive input validation** in `net/xfrm/xfrm_user.c:verify_replay()`: ### For Non-ESN Mode (32-bit sequences): - **Output SA**: Validates `rs->oseq != U32_MAX` (lines 189-195) - **Input SA**: Validates `rs->seq != U32_MAX` (lines 223-228) ### For ESN Mode (64-bit sequences): - **Output SA**: Validates `!(rs->oseq == U32_MAX && rs->oseq_hi == U32_MAX)` (lines 196-202) - **Input SA**: Validates `!(rs->seq == U32_MAX && rs->seq_hi == U32_MAX)` (lines 230-236) ## Why This Should Be Backported ### 1. **Fixes User- Visible Bug** This prevents user misconfiguration from causing immediate IPsec traffic failure, which is a critical networking bug. ### 2. **Small, Contained Fix** - **Single file modified**: `net/xfrm/xfrm_user.c` - **Only 42 insertions, 10 deletions** - **Pure input validation** - no algorithmic or architectural changes - **Low regression risk** - only rejects previously invalid configurations ### 3. **Benefits Hardware Offload** The commit message explicitly mentions this helps "packet offloads implementations which need to configure HW with next SEQ to send." This is increasingly important as IPsec hardware offload becomes more common. ### 4. **Follows Historical Pattern** Looking at similar commits in the reference examples: - **Similar Commit #1** (Status: NO) - Only validates ESN vs non-ESN mode consistency - **Current commit** - **More comprehensive**, validates against the problematic U32_MAX boundary that causes actual packet drops - **Similar Commits #3-5** (Status: YES) - All fix sequence number handling bugs that cause packet drops/corruption ### 5. **Clear Error Messages** The fix provides descriptive error messages via `NL_SET_ERR_MSG()`, improving debuggability for users. ### 6. **Builds on Previous Work** This extends the validation framework established in commit `e3aa43a50a64` ("xfrm: prevent high SEQ input in non-ESN mode"), showing this is part of ongoing hardening efforts. ## Risk Assessment **Minimal Risk**: - **No functional changes** to existing working configurations - **Only affects invalid configurations** that would cause problems anyway - **Well-tested code path** (input validation in userspace interface) - **Conservative approach** - rejects edge case that causes guaranteed failure The fix prevents a **user-triggerable traffic failure condition** with minimal code changes and no risk to existing working setups. This perfectly fits stable tree criteria: important bug fix, low risk, contained scope.
net/xfrm/xfrm_user.c | 52 +++++++++++++++++++++++++++++++++++--------- 1 file changed, 42 insertions(+), 10 deletions(-)
diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c index 784a2d124749f..614b58cb26ab7 100644 --- a/net/xfrm/xfrm_user.c +++ b/net/xfrm/xfrm_user.c @@ -178,11 +178,27 @@ static inline int verify_replay(struct xfrm_usersa_info *p, "Replay seq and seq_hi should be 0 for output SA"); return -EINVAL; } - if (rs->oseq_hi && !(p->flags & XFRM_STATE_ESN)) { - NL_SET_ERR_MSG( - extack, - "Replay oseq_hi should be 0 in non-ESN mode for output SA"); - return -EINVAL; + + if (!(p->flags & XFRM_STATE_ESN)) { + if (rs->oseq_hi) { + NL_SET_ERR_MSG( + extack, + "Replay oseq_hi should be 0 in non-ESN mode for output SA"); + return -EINVAL; + } + if (rs->oseq == U32_MAX) { + NL_SET_ERR_MSG( + extack, + "Replay oseq should be less than 0xFFFFFFFF in non-ESN mode for output SA"); + return -EINVAL; + } + } else { + if (rs->oseq == U32_MAX && rs->oseq_hi == U32_MAX) { + NL_SET_ERR_MSG( + extack, + "Replay oseq and oseq_hi should be less than 0xFFFFFFFF for output SA"); + return -EINVAL; + } } if (rs->bmp_len) { NL_SET_ERR_MSG(extack, "Replay bmp_len should 0 for output SA"); @@ -196,11 +212,27 @@ static inline int verify_replay(struct xfrm_usersa_info *p, "Replay oseq and oseq_hi should be 0 for input SA"); return -EINVAL; } - if (rs->seq_hi && !(p->flags & XFRM_STATE_ESN)) { - NL_SET_ERR_MSG( - extack, - "Replay seq_hi should be 0 in non-ESN mode for input SA"); - return -EINVAL; + if (!(p->flags & XFRM_STATE_ESN)) { + if (rs->seq_hi) { + NL_SET_ERR_MSG( + extack, + "Replay seq_hi should be 0 in non-ESN mode for input SA"); + return -EINVAL; + } + + if (rs->seq == U32_MAX) { + NL_SET_ERR_MSG( + extack, + "Replay seq should be less than 0xFFFFFFFF in non-ESN mode for input SA"); + return -EINVAL; + } + } else { + if (rs->seq == U32_MAX && rs->seq_hi == U32_MAX) { + NL_SET_ERR_MSG( + extack, + "Replay seq and seq_hi should be less than 0xFFFFFFFF for input SA"); + return -EINVAL; + } } }