Message ID | 20240304230845.14934-1-jesse.brandeburg@intel.com (mailing list archive) |
---|---|
State | Changes Requested |
Delegated to: | Netdev Maintainers |
Headers | show |
Series | [iwl-net,v1] ice: fix bug with suspend and rebuild | expand |
Dear Jesse, Thank you for your patch. Am 05.03.24 um 00:08 schrieb Jesse Brandeburg: > The ice driver would previously panic during suspend. This is caused > from the driver *only* calling the ice_vsi_free_q_vectors() function by > itself, when it is suspending. Since commit b3e7b3a6ee92 ("ice: prevent > NULL pointer deref during reload") the driver has zeroed out > num_q_vectors, and only restored it in ice_vsi_cfg_def(). > > This further causes the ice_rebuild() function to allocate a zero length > buffer, after which num_q_vectors is updated, and then the new value of > num_q_vectors is used to index into the zero length buffer, which > corrupts memory. […] For the commit message summary I suggest to be more specific. Maybe: ice: Fix memory corruption with suspend and rebuild ice: Avoid 0-length buffer to fix memory corruption with suspend/rebuild Reviewed-by: Paul Menzel <pmenzel@molgen.mpg.de> Kind regards, Paul
On Mon, Mar 04, 2024 at 03:08:44PM -0800, Jesse Brandeburg wrote: > The ice driver would previously panic during suspend. This is caused > from the driver *only* calling the ice_vsi_free_q_vectors() function by > itself, when it is suspending. Since commit b3e7b3a6ee92 ("ice: prevent > NULL pointer deref during reload") the driver has zeroed out > num_q_vectors, and only restored it in ice_vsi_cfg_def(). > > This further causes the ice_rebuild() function to allocate a zero length > buffer, after which num_q_vectors is updated, and then the new value of > num_q_vectors is used to index into the zero length buffer, which > corrupts memory. > > The fix entails making sure all the code referencing num_q_vectors only > does so after it has been reset via ice_vsi_cfg_def(). > > I didn't perform a full bisect, but I was able to test against 6.1.77 > kernel and that ice driver works fine for suspend/resume with no panic, > so sometime since then, this problem was introduced. > > Also clean up an un-needed init of a local variable in the function > being modified. > > PANIC from 6.8.0-rc1: > > [1026674.915596] PM: suspend exit > [1026675.664697] ice 0000:17:00.1: PTP reset successful > [1026675.664707] ice 0000:17:00.1: 2755 msecs passed between update to cached PHC time > [1026675.667660] ice 0000:b1:00.0: PTP reset successful > [1026675.675944] ice 0000:b1:00.0: 2832 msecs passed between update to cached PHC time > [1026677.137733] ixgbe 0000:31:00.0 ens787: NIC Link is Up 1 Gbps, Flow Control: None > [1026677.190201] BUG: kernel NULL pointer dereference, address: 0000000000000010 > [1026677.192753] ice 0000:17:00.0: PTP reset successful > [1026677.192764] ice 0000:17:00.0: 4548 msecs passed between update to cached PHC time > [1026677.197928] #PF: supervisor read access in kernel mode > [1026677.197933] #PF: error_code(0x0000) - not-present page > [1026677.197937] PGD 1557a7067 P4D 0 > [1026677.212133] ice 0000:b1:00.1: PTP reset successful > [1026677.212143] ice 0000:b1:00.1: 4344 msecs passed between update to cached PHC time > [1026677.212575] > [1026677.243142] Oops: 0000 [#1] PREEMPT SMP NOPTI > [1026677.247918] CPU: 23 PID: 42790 Comm: kworker/23:0 Kdump: loaded Tainted: G W 6.8.0-rc1+ #1 > [1026677.257989] Hardware name: Intel Corporation M50CYP2SBSTD/M50CYP2SBSTD, BIOS SE5C620.86B.01.01.0005.2202160810 02/16/2022 > [1026677.269367] Workqueue: ice ice_service_task [ice] > [1026677.274592] RIP: 0010:ice_vsi_rebuild_set_coalesce+0x130/0x1e0 [ice] > [1026677.281421] Code: 0f 84 3a ff ff ff 41 0f b7 74 ec 02 66 89 b0 22 02 00 00 81 e6 ff 1f 00 00 e8 ec fd ff ff e9 35 ff ff ff 48 8b 43 30 49 63 ed <41> 0f b7 34 24 41 83 c5 01 48 8b 3c e8 66 89 b7 aa 02 00 00 81 e6 > [1026677.300877] RSP: 0018:ff3be62a6399bcc0 EFLAGS: 00010202 > [1026677.306556] RAX: ff28691e28980828 RBX: ff28691e41099828 RCX: 0000000000188000 > [1026677.314148] RDX: 0000000000000000 RSI: 0000000000000010 RDI: ff28691e41099828 > [1026677.321730] RBP: 0000000000000000 R08: 0000000000000000 R09: 0000000000000000 > [1026677.329311] R10: 0000000000000007 R11: ffffffffffffffc0 R12: 0000000000000010 > [1026677.336896] R13: 0000000000000000 R14: 0000000000000000 R15: ff28691e0eaa81a0 > [1026677.344472] FS: 0000000000000000(0000) GS:ff28693cbffc0000(0000) knlGS:0000000000000000 > [1026677.353000] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 > [1026677.359195] CR2: 0000000000000010 CR3: 0000000128df4001 CR4: 0000000000771ef0 > [1026677.366779] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 > [1026677.374369] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 > [1026677.381952] PKRU: 55555554 > [1026677.385116] Call Trace: > [1026677.388023] <TASK> > [1026677.390589] ? __die+0x20/0x70 > [1026677.394105] ? page_fault_oops+0x82/0x160 > [1026677.398576] ? do_user_addr_fault+0x65/0x6a0 > [1026677.403307] ? exc_page_fault+0x6a/0x150 > [1026677.407694] ? asm_exc_page_fault+0x22/0x30 > [1026677.412349] ? ice_vsi_rebuild_set_coalesce+0x130/0x1e0 [ice] > [1026677.418614] ice_vsi_rebuild+0x34b/0x3c0 [ice] > [1026677.423583] ice_vsi_rebuild_by_type+0x76/0x180 [ice] > [1026677.429147] ice_rebuild+0x18b/0x520 [ice] > [1026677.433746] ? delay_tsc+0x8f/0xc0 > [1026677.437630] ice_do_reset+0xa3/0x190 [ice] > [1026677.442231] ice_service_task+0x26/0x440 [ice] > [1026677.447180] process_one_work+0x174/0x340 > [1026677.451669] worker_thread+0x27e/0x390 > [1026677.455890] ? __pfx_worker_thread+0x10/0x10 > [1026677.460627] kthread+0xee/0x120 > [1026677.464235] ? __pfx_kthread+0x10/0x10 > [1026677.468445] ret_from_fork+0x2d/0x50 > [1026677.472476] ? __pfx_kthread+0x10/0x10 > [1026677.476671] ret_from_fork_asm+0x1b/0x30 > [1026677.481050] </TASK> > > Fixes: b3e7b3a6ee92 ("ice: prevent NULL pointer deref during reload") > Reported-by: Robert Elliott <elliott@hpe.com> > Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> > --- > drivers/net/ethernet/intel/ice/ice_lib.c | 16 ++++++++-------- > 1 file changed, 8 insertions(+), 8 deletions(-) > > diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c > index 097bf8fd6bf0..0f5a92a6b1e6 100644 > --- a/drivers/net/ethernet/intel/ice/ice_lib.c > +++ b/drivers/net/ethernet/intel/ice/ice_lib.c > @@ -3238,7 +3238,7 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, u32 vsi_flags) > { > struct ice_vsi_cfg_params params = {}; > struct ice_coalesce_stored *coalesce; > - int prev_num_q_vectors = 0; > + int prev_num_q_vectors; > struct ice_pf *pf; > int ret; > > @@ -3252,13 +3252,6 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, u32 vsi_flags) > if (WARN_ON(vsi->type == ICE_VSI_VF && !vsi->vf)) > return -EINVAL; > > - coalesce = kcalloc(vsi->num_q_vectors, > - sizeof(struct ice_coalesce_stored), GFP_KERNEL); > - if (!coalesce) > - return -ENOMEM; > - > - prev_num_q_vectors = ice_vsi_rebuild_get_coalesce(vsi, coalesce); > - > ret = ice_vsi_realloc_stat_arrays(vsi); > if (ret) > goto err_vsi_cfg; > @@ -3268,6 +3261,13 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, u32 vsi_flags) > if (ret) > goto err_vsi_cfg; Hi Jesse, the label above will result in a call to kfree(coalesce). However, coalesce is now uninitialised until the following line executes. > > + coalesce = kcalloc(vsi->num_q_vectors, > + sizeof(struct ice_coalesce_stored), GFP_KERNEL); > + if (!coalesce) > + return -ENOMEM; > + > + prev_num_q_vectors = ice_vsi_rebuild_get_coalesce(vsi, coalesce); > + > ret = ice_vsi_cfg_tc_lan(pf, vsi); > if (ret) { > if (vsi_flags & ICE_VSI_FLAG_INIT) { > > base-commit: 6923134fc6b62d7909169b3ad913ab72ee04233a
On 3/5/2024 5:17 AM, Simon Horman wrote: --- a/drivers/net/ethernet/intel/ice/ice_lib.c >> +++ b/drivers/net/ethernet/intel/ice/ice_lib.c >> @@ -3238,7 +3238,7 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, u32 vsi_flags) >> { >> struct ice_vsi_cfg_params params = {}; >> struct ice_coalesce_stored *coalesce; >> - int prev_num_q_vectors = 0; >> + int prev_num_q_vectors; >> struct ice_pf *pf; >> int ret; >> >> @@ -3252,13 +3252,6 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, u32 vsi_flags) >> if (WARN_ON(vsi->type == ICE_VSI_VF && !vsi->vf)) >> return -EINVAL; >> >> - coalesce = kcalloc(vsi->num_q_vectors, >> - sizeof(struct ice_coalesce_stored), GFP_KERNEL); >> - if (!coalesce) >> - return -ENOMEM; >> - >> - prev_num_q_vectors = ice_vsi_rebuild_get_coalesce(vsi, coalesce); >> - >> ret = ice_vsi_realloc_stat_arrays(vsi); >> if (ret) >> goto err_vsi_cfg; >> @@ -3268,6 +3261,13 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, u32 vsi_flags) >> if (ret) >> goto err_vsi_cfg; > > Hi Jesse, > > the label above will result in a call to kfree(coalesce). > However, coalesce is now uninitialised until the following line executes. Thanks Simon, you're right! And it figures, The internal static analysis runs caught this only after I sent the patch to the list. One fix is to NULL the initialization of coalesce, which solves the problem with a one-liner, but I'll look more today at whether I should just move the label or something else.
On Mon, Mar 04, 2024 at 03:08:44PM -0800, Jesse Brandeburg wrote: > The ice driver would previously panic during suspend. This is caused > from the driver *only* calling the ice_vsi_free_q_vectors() function by > itself, when it is suspending. Since commit b3e7b3a6ee92 ("ice: prevent > NULL pointer deref during reload") the driver has zeroed out > num_q_vectors, and only restored it in ice_vsi_cfg_def(). > > This further causes the ice_rebuild() function to allocate a zero length > buffer, after which num_q_vectors is updated, and then the new value of > num_q_vectors is used to index into the zero length buffer, which > corrupts memory. > > The fix entails making sure all the code referencing num_q_vectors only > does so after it has been reset via ice_vsi_cfg_def(). > > I didn't perform a full bisect, but I was able to test against 6.1.77 > kernel and that ice driver works fine for suspend/resume with no panic, > so sometime since then, this problem was introduced. > > Also clean up an un-needed init of a local variable in the function > being modified. > > PANIC from 6.8.0-rc1: > > [1026674.915596] PM: suspend exit > [1026675.664697] ice 0000:17:00.1: PTP reset successful > [1026675.664707] ice 0000:17:00.1: 2755 msecs passed between update to cached PHC time > [1026675.667660] ice 0000:b1:00.0: PTP reset successful > [1026675.675944] ice 0000:b1:00.0: 2832 msecs passed between update to cached PHC time > [1026677.137733] ixgbe 0000:31:00.0 ens787: NIC Link is Up 1 Gbps, Flow Control: None > [1026677.190201] BUG: kernel NULL pointer dereference, address: 0000000000000010 > [1026677.192753] ice 0000:17:00.0: PTP reset successful > [1026677.192764] ice 0000:17:00.0: 4548 msecs passed between update to cached PHC time > [1026677.197928] #PF: supervisor read access in kernel mode > [1026677.197933] #PF: error_code(0x0000) - not-present page > [1026677.197937] PGD 1557a7067 P4D 0 > [1026677.212133] ice 0000:b1:00.1: PTP reset successful > [1026677.212143] ice 0000:b1:00.1: 4344 msecs passed between update to cached PHC time > [1026677.212575] > [1026677.243142] Oops: 0000 [#1] PREEMPT SMP NOPTI > [1026677.247918] CPU: 23 PID: 42790 Comm: kworker/23:0 Kdump: loaded Tainted: G W 6.8.0-rc1+ #1 > [1026677.257989] Hardware name: Intel Corporation M50CYP2SBSTD/M50CYP2SBSTD, BIOS SE5C620.86B.01.01.0005.2202160810 02/16/2022 > [1026677.269367] Workqueue: ice ice_service_task [ice] > [1026677.274592] RIP: 0010:ice_vsi_rebuild_set_coalesce+0x130/0x1e0 [ice] > [1026677.281421] Code: 0f 84 3a ff ff ff 41 0f b7 74 ec 02 66 89 b0 22 02 00 00 81 e6 ff 1f 00 00 e8 ec fd ff ff e9 35 ff ff ff 48 8b 43 30 49 63 ed <41> 0f b7 34 24 41 83 c5 01 48 8b 3c e8 66 89 b7 aa 02 00 00 81 e6 > [1026677.300877] RSP: 0018:ff3be62a6399bcc0 EFLAGS: 00010202 > [1026677.306556] RAX: ff28691e28980828 RBX: ff28691e41099828 RCX: 0000000000188000 > [1026677.314148] RDX: 0000000000000000 RSI: 0000000000000010 RDI: ff28691e41099828 > [1026677.321730] RBP: 0000000000000000 R08: 0000000000000000 R09: 0000000000000000 > [1026677.329311] R10: 0000000000000007 R11: ffffffffffffffc0 R12: 0000000000000010 > [1026677.336896] R13: 0000000000000000 R14: 0000000000000000 R15: ff28691e0eaa81a0 > [1026677.344472] FS: 0000000000000000(0000) GS:ff28693cbffc0000(0000) knlGS:0000000000000000 > [1026677.353000] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 > [1026677.359195] CR2: 0000000000000010 CR3: 0000000128df4001 CR4: 0000000000771ef0 > [1026677.366779] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 > [1026677.374369] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 > [1026677.381952] PKRU: 55555554 > [1026677.385116] Call Trace: > [1026677.388023] <TASK> > [1026677.390589] ? __die+0x20/0x70 > [1026677.394105] ? page_fault_oops+0x82/0x160 > [1026677.398576] ? do_user_addr_fault+0x65/0x6a0 > [1026677.403307] ? exc_page_fault+0x6a/0x150 > [1026677.407694] ? asm_exc_page_fault+0x22/0x30 > [1026677.412349] ? ice_vsi_rebuild_set_coalesce+0x130/0x1e0 [ice] > [1026677.418614] ice_vsi_rebuild+0x34b/0x3c0 [ice] > [1026677.423583] ice_vsi_rebuild_by_type+0x76/0x180 [ice] > [1026677.429147] ice_rebuild+0x18b/0x520 [ice] > [1026677.433746] ? delay_tsc+0x8f/0xc0 > [1026677.437630] ice_do_reset+0xa3/0x190 [ice] > [1026677.442231] ice_service_task+0x26/0x440 [ice] > [1026677.447180] process_one_work+0x174/0x340 > [1026677.451669] worker_thread+0x27e/0x390 > [1026677.455890] ? __pfx_worker_thread+0x10/0x10 > [1026677.460627] kthread+0xee/0x120 > [1026677.464235] ? __pfx_kthread+0x10/0x10 > [1026677.468445] ret_from_fork+0x2d/0x50 > [1026677.472476] ? __pfx_kthread+0x10/0x10 > [1026677.476671] ret_from_fork_asm+0x1b/0x30 > [1026677.481050] </TASK> > > Fixes: b3e7b3a6ee92 ("ice: prevent NULL pointer deref during reload") > Reported-by: Robert Elliott <elliott@hpe.com> > Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com> Well, that refactor of config path introduced lots of issues. Could validation folks include a short list of tests they tried out against tested patch? > --- > drivers/net/ethernet/intel/ice/ice_lib.c | 16 ++++++++-------- > 1 file changed, 8 insertions(+), 8 deletions(-) > > diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c > index 097bf8fd6bf0..0f5a92a6b1e6 100644 > --- a/drivers/net/ethernet/intel/ice/ice_lib.c > +++ b/drivers/net/ethernet/intel/ice/ice_lib.c > @@ -3238,7 +3238,7 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, u32 vsi_flags) > { > struct ice_vsi_cfg_params params = {}; > struct ice_coalesce_stored *coalesce; > - int prev_num_q_vectors = 0; > + int prev_num_q_vectors; > struct ice_pf *pf; > int ret; > > @@ -3252,13 +3252,6 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, u32 vsi_flags) > if (WARN_ON(vsi->type == ICE_VSI_VF && !vsi->vf)) > return -EINVAL; > > - coalesce = kcalloc(vsi->num_q_vectors, > - sizeof(struct ice_coalesce_stored), GFP_KERNEL); > - if (!coalesce) > - return -ENOMEM; > - > - prev_num_q_vectors = ice_vsi_rebuild_get_coalesce(vsi, coalesce); > - > ret = ice_vsi_realloc_stat_arrays(vsi); > if (ret) > goto err_vsi_cfg; > @@ -3268,6 +3261,13 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, u32 vsi_flags) > if (ret) > goto err_vsi_cfg; > > + coalesce = kcalloc(vsi->num_q_vectors, > + sizeof(struct ice_coalesce_stored), GFP_KERNEL); > + if (!coalesce) > + return -ENOMEM; > + > + prev_num_q_vectors = ice_vsi_rebuild_get_coalesce(vsi, coalesce); > + > ret = ice_vsi_cfg_tc_lan(pf, vsi); > if (ret) { > if (vsi_flags & ICE_VSI_FLAG_INIT) { > > base-commit: 6923134fc6b62d7909169b3ad913ab72ee04233a > -- > 2.39.3 > >
On Wed, Mar 06, 2024 at 05:46:02PM +0100, Maciej Fijalkowski wrote: > On Mon, Mar 04, 2024 at 03:08:44PM -0800, Jesse Brandeburg wrote: > > The ice driver would previously panic during suspend. This is caused > > from the driver *only* calling the ice_vsi_free_q_vectors() function by > > itself, when it is suspending. Since commit b3e7b3a6ee92 ("ice: prevent > > NULL pointer deref during reload") the driver has zeroed out > > num_q_vectors, and only restored it in ice_vsi_cfg_def(). > > > > This further causes the ice_rebuild() function to allocate a zero length > > buffer, after which num_q_vectors is updated, and then the new value of > > num_q_vectors is used to index into the zero length buffer, which > > corrupts memory. > > > > The fix entails making sure all the code referencing num_q_vectors only > > does so after it has been reset via ice_vsi_cfg_def(). > > > > I didn't perform a full bisect, but I was able to test against 6.1.77 > > kernel and that ice driver works fine for suspend/resume with no panic, > > so sometime since then, this problem was introduced. > > > > Also clean up an un-needed init of a local variable in the function > > being modified. > > > > PANIC from 6.8.0-rc1: > > > > [1026674.915596] PM: suspend exit > > [1026675.664697] ice 0000:17:00.1: PTP reset successful > > [1026675.664707] ice 0000:17:00.1: 2755 msecs passed between update to cached PHC time > > [1026675.667660] ice 0000:b1:00.0: PTP reset successful > > [1026675.675944] ice 0000:b1:00.0: 2832 msecs passed between update to cached PHC time > > [1026677.137733] ixgbe 0000:31:00.0 ens787: NIC Link is Up 1 Gbps, Flow Control: None > > [1026677.190201] BUG: kernel NULL pointer dereference, address: 0000000000000010 > > [1026677.192753] ice 0000:17:00.0: PTP reset successful > > [1026677.192764] ice 0000:17:00.0: 4548 msecs passed between update to cached PHC time > > [1026677.197928] #PF: supervisor read access in kernel mode > > [1026677.197933] #PF: error_code(0x0000) - not-present page > > [1026677.197937] PGD 1557a7067 P4D 0 > > [1026677.212133] ice 0000:b1:00.1: PTP reset successful > > [1026677.212143] ice 0000:b1:00.1: 4344 msecs passed between update to cached PHC time > > [1026677.212575] > > [1026677.243142] Oops: 0000 [#1] PREEMPT SMP NOPTI > > [1026677.247918] CPU: 23 PID: 42790 Comm: kworker/23:0 Kdump: loaded Tainted: G W 6.8.0-rc1+ #1 > > [1026677.257989] Hardware name: Intel Corporation M50CYP2SBSTD/M50CYP2SBSTD, BIOS SE5C620.86B.01.01.0005.2202160810 02/16/2022 > > [1026677.269367] Workqueue: ice ice_service_task [ice] > > [1026677.274592] RIP: 0010:ice_vsi_rebuild_set_coalesce+0x130/0x1e0 [ice] > > [1026677.281421] Code: 0f 84 3a ff ff ff 41 0f b7 74 ec 02 66 89 b0 22 02 00 00 81 e6 ff 1f 00 00 e8 ec fd ff ff e9 35 ff ff ff 48 8b 43 30 49 63 ed <41> 0f b7 34 24 41 83 c5 01 48 8b 3c e8 66 89 b7 aa 02 00 00 81 e6 > > [1026677.300877] RSP: 0018:ff3be62a6399bcc0 EFLAGS: 00010202 > > [1026677.306556] RAX: ff28691e28980828 RBX: ff28691e41099828 RCX: 0000000000188000 > > [1026677.314148] RDX: 0000000000000000 RSI: 0000000000000010 RDI: ff28691e41099828 > > [1026677.321730] RBP: 0000000000000000 R08: 0000000000000000 R09: 0000000000000000 > > [1026677.329311] R10: 0000000000000007 R11: ffffffffffffffc0 R12: 0000000000000010 > > [1026677.336896] R13: 0000000000000000 R14: 0000000000000000 R15: ff28691e0eaa81a0 > > [1026677.344472] FS: 0000000000000000(0000) GS:ff28693cbffc0000(0000) knlGS:0000000000000000 > > [1026677.353000] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 > > [1026677.359195] CR2: 0000000000000010 CR3: 0000000128df4001 CR4: 0000000000771ef0 > > [1026677.366779] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 > > [1026677.374369] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 > > [1026677.381952] PKRU: 55555554 > > [1026677.385116] Call Trace: > > [1026677.388023] <TASK> > > [1026677.390589] ? __die+0x20/0x70 > > [1026677.394105] ? page_fault_oops+0x82/0x160 > > [1026677.398576] ? do_user_addr_fault+0x65/0x6a0 > > [1026677.403307] ? exc_page_fault+0x6a/0x150 > > [1026677.407694] ? asm_exc_page_fault+0x22/0x30 > > [1026677.412349] ? ice_vsi_rebuild_set_coalesce+0x130/0x1e0 [ice] > > [1026677.418614] ice_vsi_rebuild+0x34b/0x3c0 [ice] > > [1026677.423583] ice_vsi_rebuild_by_type+0x76/0x180 [ice] > > [1026677.429147] ice_rebuild+0x18b/0x520 [ice] > > [1026677.433746] ? delay_tsc+0x8f/0xc0 > > [1026677.437630] ice_do_reset+0xa3/0x190 [ice] > > [1026677.442231] ice_service_task+0x26/0x440 [ice] > > [1026677.447180] process_one_work+0x174/0x340 > > [1026677.451669] worker_thread+0x27e/0x390 > > [1026677.455890] ? __pfx_worker_thread+0x10/0x10 > > [1026677.460627] kthread+0xee/0x120 > > [1026677.464235] ? __pfx_kthread+0x10/0x10 > > [1026677.468445] ret_from_fork+0x2d/0x50 > > [1026677.472476] ? __pfx_kthread+0x10/0x10 > > [1026677.476671] ret_from_fork_asm+0x1b/0x30 > > [1026677.481050] </TASK> > > > > Fixes: b3e7b3a6ee92 ("ice: prevent NULL pointer deref during reload") > > Reported-by: Robert Elliott <elliott@hpe.com> > > Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> > > Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com> > > Well, that refactor of config path introduced lots of issues. Could > validation folks include a short list of tests they tried out against > tested patch? Sorry, I got confused and now I saw the same thing Simon pointed out. > > > --- > > drivers/net/ethernet/intel/ice/ice_lib.c | 16 ++++++++-------- > > 1 file changed, 8 insertions(+), 8 deletions(-) > > > > diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c > > index 097bf8fd6bf0..0f5a92a6b1e6 100644 > > --- a/drivers/net/ethernet/intel/ice/ice_lib.c > > +++ b/drivers/net/ethernet/intel/ice/ice_lib.c > > @@ -3238,7 +3238,7 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, u32 vsi_flags) > > { > > struct ice_vsi_cfg_params params = {}; > > struct ice_coalesce_stored *coalesce; struct ice_coalesce_stored *coalesce __free(kfree); ? and drop explicit kfree()s altogether? > > - int prev_num_q_vectors = 0; > > + int prev_num_q_vectors; > > struct ice_pf *pf; > > int ret; > > > > @@ -3252,13 +3252,6 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, u32 vsi_flags) > > if (WARN_ON(vsi->type == ICE_VSI_VF && !vsi->vf)) > > return -EINVAL; > > > > - coalesce = kcalloc(vsi->num_q_vectors, > > - sizeof(struct ice_coalesce_stored), GFP_KERNEL); > > - if (!coalesce) > > - return -ENOMEM; > > - > > - prev_num_q_vectors = ice_vsi_rebuild_get_coalesce(vsi, coalesce); > > - > > ret = ice_vsi_realloc_stat_arrays(vsi); > > if (ret) > > goto err_vsi_cfg; > > @@ -3268,6 +3261,13 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, u32 vsi_flags) > > if (ret) > > goto err_vsi_cfg; > > > > + coalesce = kcalloc(vsi->num_q_vectors, > > + sizeof(struct ice_coalesce_stored), GFP_KERNEL); > > + if (!coalesce) > > + return -ENOMEM; > > + > > + prev_num_q_vectors = ice_vsi_rebuild_get_coalesce(vsi, coalesce); > > + > > ret = ice_vsi_cfg_tc_lan(pf, vsi); > > if (ret) { > > if (vsi_flags & ICE_VSI_FLAG_INIT) { > > > > base-commit: 6923134fc6b62d7909169b3ad913ab72ee04233a > > -- > > 2.39.3 > > > > >
On 3/6/2024 8:56 AM, Maciej Fijalkowski wrote: >>> Fixes: b3e7b3a6ee92 ("ice: prevent NULL pointer deref during reload") >>> Reported-by: Robert Elliott <elliott@hpe.com> >>> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> >> >> Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com> >> >> Well, that refactor of config path introduced lots of issues. Could >> validation folks include a short list of tests they tried out against >> tested patch? > > Sorry, I got confused and now I saw the same thing Simon pointed out. Thanks, NP! this thread is for v1, and v2 is already posted with the refactor just moving the CONFIG_FOO in patch 1, and then the real final version of the fix in v2 patch two which switches all the intel drivers over to the new macros. > >> >>> --- >>> drivers/net/ethernet/intel/ice/ice_lib.c | 16 ++++++++-------- >>> 1 file changed, 8 insertions(+), 8 deletions(-) >>> >>> diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c >>> index 097bf8fd6bf0..0f5a92a6b1e6 100644 >>> --- a/drivers/net/ethernet/intel/ice/ice_lib.c >>> +++ b/drivers/net/ethernet/intel/ice/ice_lib.c >>> @@ -3238,7 +3238,7 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, u32 vsi_flags) >>> { >>> struct ice_vsi_cfg_params params = {}; >>> struct ice_coalesce_stored *coalesce; > > struct ice_coalesce_stored *coalesce __free(kfree); Yes, but not in a "fix" patch for -net. I figure this is more of a refactor kind of thing, so will not include it in this change. > > ? > > and drop explicit kfree()s altogether? >
On 3/6/2024 5:33 PM, Jesse Brandeburg wrote: > On 3/6/2024 8:56 AM, Maciej Fijalkowski wrote: >>>> Fixes: b3e7b3a6ee92 ("ice: prevent NULL pointer deref during reload") >>>> Reported-by: Robert Elliott <elliott@hpe.com> >>>> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> >>> >>> Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com> >>> >>> Well, that refactor of config path introduced lots of issues. Could >>> validation folks include a short list of tests they tried out against >>> tested patch? That's a good question which I'll follow up on with our team, but please don't let it block anything for this patch. >> >> Sorry, I got confused and now I saw the same thing Simon pointed out. > > Thanks, NP! this thread is for v1, and v2 is already posted with the > refactor just moving the CONFIG_FOO in patch 1, and then the real final > version of the fix in v2 patch two which switches all the intel drivers > over to the new macros. oops! FYI v2 of this change was posted at https://lore.kernel.org/netdev/20240305230204.448724-1-jesse.brandeburg@intel.com/ Thanks for the review!
On Wed, Mar 06, 2024 at 05:41:10PM -0800, Jesse Brandeburg wrote: > On 3/6/2024 5:33 PM, Jesse Brandeburg wrote: > > On 3/6/2024 8:56 AM, Maciej Fijalkowski wrote: > > > > > Fixes: b3e7b3a6ee92 ("ice: prevent NULL pointer deref during reload") > > > > > Reported-by: Robert Elliott <elliott@hpe.com> > > > > > Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> > > > > > > > > Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com> > > > > > > > > Well, that refactor of config path introduced lots of issues. Could > > > > validation folks include a short list of tests they tried out against > > > > tested patch? > > That's a good question which I'll follow up on with our team, but please > don't let it block anything for this patch. Blocking was not my intent at all, just raising a concern regarding this ever-lasting issue. > > > > > > > Sorry, I got confused and now I saw the same thing Simon pointed out. > > > > Thanks, NP! this thread is for v1, and v2 is already posted with the > > refactor just moving the CONFIG_FOO in patch 1, and then the real final > > version of the fix in v2 patch two which switches all the intel drivers > > over to the new macros. > > oops! > > FYI v2 of this change was posted at https://lore.kernel.org/netdev/20240305230204.448724-1-jesse.brandeburg@intel.com/ > > Thanks for the review! >
On Wed, Mar 06, 2024 at 05:33:50PM -0800, Jesse Brandeburg wrote: > On 3/6/2024 8:56 AM, Maciej Fijalkowski wrote: > > > > Fixes: b3e7b3a6ee92 ("ice: prevent NULL pointer deref during reload") > > > > Reported-by: Robert Elliott <elliott@hpe.com> > > > > Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> > > > > > > Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com> > > > > > > Well, that refactor of config path introduced lots of issues. Could > > > validation folks include a short list of tests they tried out against > > > tested patch? > > > > Sorry, I got confused and now I saw the same thing Simon pointed out. > > Thanks, NP! this thread is for v1, and v2 is already posted with the > refactor just moving the CONFIG_FOO in patch 1, and then the real final > version of the fix in v2 patch two which switches all the intel drivers over > to the new macros. > > > > > > > > > > --- > > > > drivers/net/ethernet/intel/ice/ice_lib.c | 16 ++++++++-------- > > > > 1 file changed, 8 insertions(+), 8 deletions(-) > > > > > > > > diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c > > > > index 097bf8fd6bf0..0f5a92a6b1e6 100644 > > > > --- a/drivers/net/ethernet/intel/ice/ice_lib.c > > > > +++ b/drivers/net/ethernet/intel/ice/ice_lib.c > > > > @@ -3238,7 +3238,7 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, u32 vsi_flags) > > > > { > > > > struct ice_vsi_cfg_params params = {}; > > > > struct ice_coalesce_stored *coalesce; > > > > struct ice_coalesce_stored *coalesce __free(kfree); > > Yes, but not in a "fix" patch for -net. I figure this is more of a refactor > kind of thing, so will not include it in this change. I suppose that's because older kernels might not have this construct backported, so yeah your v2 is good as-is. > > > > > ? > > > > and drop explicit kfree()s altogether? > > > >
diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c index 097bf8fd6bf0..0f5a92a6b1e6 100644 --- a/drivers/net/ethernet/intel/ice/ice_lib.c +++ b/drivers/net/ethernet/intel/ice/ice_lib.c @@ -3238,7 +3238,7 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, u32 vsi_flags) { struct ice_vsi_cfg_params params = {}; struct ice_coalesce_stored *coalesce; - int prev_num_q_vectors = 0; + int prev_num_q_vectors; struct ice_pf *pf; int ret; @@ -3252,13 +3252,6 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, u32 vsi_flags) if (WARN_ON(vsi->type == ICE_VSI_VF && !vsi->vf)) return -EINVAL; - coalesce = kcalloc(vsi->num_q_vectors, - sizeof(struct ice_coalesce_stored), GFP_KERNEL); - if (!coalesce) - return -ENOMEM; - - prev_num_q_vectors = ice_vsi_rebuild_get_coalesce(vsi, coalesce); - ret = ice_vsi_realloc_stat_arrays(vsi); if (ret) goto err_vsi_cfg; @@ -3268,6 +3261,13 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, u32 vsi_flags) if (ret) goto err_vsi_cfg; + coalesce = kcalloc(vsi->num_q_vectors, + sizeof(struct ice_coalesce_stored), GFP_KERNEL); + if (!coalesce) + return -ENOMEM; + + prev_num_q_vectors = ice_vsi_rebuild_get_coalesce(vsi, coalesce); + ret = ice_vsi_cfg_tc_lan(pf, vsi); if (ret) { if (vsi_flags & ICE_VSI_FLAG_INIT) {
The ice driver would previously panic during suspend. This is caused from the driver *only* calling the ice_vsi_free_q_vectors() function by itself, when it is suspending. Since commit b3e7b3a6ee92 ("ice: prevent NULL pointer deref during reload") the driver has zeroed out num_q_vectors, and only restored it in ice_vsi_cfg_def(). This further causes the ice_rebuild() function to allocate a zero length buffer, after which num_q_vectors is updated, and then the new value of num_q_vectors is used to index into the zero length buffer, which corrupts memory. The fix entails making sure all the code referencing num_q_vectors only does so after it has been reset via ice_vsi_cfg_def(). I didn't perform a full bisect, but I was able to test against 6.1.77 kernel and that ice driver works fine for suspend/resume with no panic, so sometime since then, this problem was introduced. Also clean up an un-needed init of a local variable in the function being modified. PANIC from 6.8.0-rc1: [1026674.915596] PM: suspend exit [1026675.664697] ice 0000:17:00.1: PTP reset successful [1026675.664707] ice 0000:17:00.1: 2755 msecs passed between update to cached PHC time [1026675.667660] ice 0000:b1:00.0: PTP reset successful [1026675.675944] ice 0000:b1:00.0: 2832 msecs passed between update to cached PHC time [1026677.137733] ixgbe 0000:31:00.0 ens787: NIC Link is Up 1 Gbps, Flow Control: None [1026677.190201] BUG: kernel NULL pointer dereference, address: 0000000000000010 [1026677.192753] ice 0000:17:00.0: PTP reset successful [1026677.192764] ice 0000:17:00.0: 4548 msecs passed between update to cached PHC time [1026677.197928] #PF: supervisor read access in kernel mode [1026677.197933] #PF: error_code(0x0000) - not-present page [1026677.197937] PGD 1557a7067 P4D 0 [1026677.212133] ice 0000:b1:00.1: PTP reset successful [1026677.212143] ice 0000:b1:00.1: 4344 msecs passed between update to cached PHC time [1026677.212575] [1026677.243142] Oops: 0000 [#1] PREEMPT SMP NOPTI [1026677.247918] CPU: 23 PID: 42790 Comm: kworker/23:0 Kdump: loaded Tainted: G W 6.8.0-rc1+ #1 [1026677.257989] Hardware name: Intel Corporation M50CYP2SBSTD/M50CYP2SBSTD, BIOS SE5C620.86B.01.01.0005.2202160810 02/16/2022 [1026677.269367] Workqueue: ice ice_service_task [ice] [1026677.274592] RIP: 0010:ice_vsi_rebuild_set_coalesce+0x130/0x1e0 [ice] [1026677.281421] Code: 0f 84 3a ff ff ff 41 0f b7 74 ec 02 66 89 b0 22 02 00 00 81 e6 ff 1f 00 00 e8 ec fd ff ff e9 35 ff ff ff 48 8b 43 30 49 63 ed <41> 0f b7 34 24 41 83 c5 01 48 8b 3c e8 66 89 b7 aa 02 00 00 81 e6 [1026677.300877] RSP: 0018:ff3be62a6399bcc0 EFLAGS: 00010202 [1026677.306556] RAX: ff28691e28980828 RBX: ff28691e41099828 RCX: 0000000000188000 [1026677.314148] RDX: 0000000000000000 RSI: 0000000000000010 RDI: ff28691e41099828 [1026677.321730] RBP: 0000000000000000 R08: 0000000000000000 R09: 0000000000000000 [1026677.329311] R10: 0000000000000007 R11: ffffffffffffffc0 R12: 0000000000000010 [1026677.336896] R13: 0000000000000000 R14: 0000000000000000 R15: ff28691e0eaa81a0 [1026677.344472] FS: 0000000000000000(0000) GS:ff28693cbffc0000(0000) knlGS:0000000000000000 [1026677.353000] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [1026677.359195] CR2: 0000000000000010 CR3: 0000000128df4001 CR4: 0000000000771ef0 [1026677.366779] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [1026677.374369] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [1026677.381952] PKRU: 55555554 [1026677.385116] Call Trace: [1026677.388023] <TASK> [1026677.390589] ? __die+0x20/0x70 [1026677.394105] ? page_fault_oops+0x82/0x160 [1026677.398576] ? do_user_addr_fault+0x65/0x6a0 [1026677.403307] ? exc_page_fault+0x6a/0x150 [1026677.407694] ? asm_exc_page_fault+0x22/0x30 [1026677.412349] ? ice_vsi_rebuild_set_coalesce+0x130/0x1e0 [ice] [1026677.418614] ice_vsi_rebuild+0x34b/0x3c0 [ice] [1026677.423583] ice_vsi_rebuild_by_type+0x76/0x180 [ice] [1026677.429147] ice_rebuild+0x18b/0x520 [ice] [1026677.433746] ? delay_tsc+0x8f/0xc0 [1026677.437630] ice_do_reset+0xa3/0x190 [ice] [1026677.442231] ice_service_task+0x26/0x440 [ice] [1026677.447180] process_one_work+0x174/0x340 [1026677.451669] worker_thread+0x27e/0x390 [1026677.455890] ? __pfx_worker_thread+0x10/0x10 [1026677.460627] kthread+0xee/0x120 [1026677.464235] ? __pfx_kthread+0x10/0x10 [1026677.468445] ret_from_fork+0x2d/0x50 [1026677.472476] ? __pfx_kthread+0x10/0x10 [1026677.476671] ret_from_fork_asm+0x1b/0x30 [1026677.481050] </TASK> Fixes: b3e7b3a6ee92 ("ice: prevent NULL pointer deref during reload") Reported-by: Robert Elliott <elliott@hpe.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> --- drivers/net/ethernet/intel/ice/ice_lib.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) base-commit: 6923134fc6b62d7909169b3ad913ab72ee04233a