Thread (8 messages) 8 messages, 3 authors, 2026-01-19

Re: [devel-ipsec] Re: [PATCH ipsec-next v2 4/4] xfrm: add XFRM_MSG_MIGRATE_STATE for single SA migration

From: Antony Antony <hidden>
Date: 2026-01-19 08:22:04
Also in: lkml, oe-kbuild, oe-kbuild-all, selinux

Hi Dan,

On Mon, Jan 19, 2026 at 08:27:25AM +0300, Dan Carpenter via Devel wrote:
Hi Antony,

kernel test robot noticed the following build warnings:

https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Antony-Antony/xfrm-remove-redundant-assignments/20260118-041031
base:   https://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec-next.git master
patch link:    https://lore.kernel.org/r/951cb30ac3866c6075bc7359d0997dbffc3ce6da.1768679141.git.antony.antony%40secunet.com
patch subject: [PATCH ipsec-next v2 4/4] xfrm: add XFRM_MSG_MIGRATE_STATE for single SA migration
config: hexagon-randconfig-r072-20260118 (https://download.01.org/0day-ci/archive/20260119/202601190605.ZVkgcUYl-lkp@intel.com/config)
compiler: clang version 22.0.0git (https://github.com/llvm/llvm-project 9b8addffa70cee5b2acc5454712d9cf78ce45710)
smatch version: v0.5.0-8985-g2614ff1a

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot [off-list ref]
| Reported-by: Dan Carpenter [off-list ref]
| Closes: https://lore.kernel.org/r/202601190605.ZVkgcUYl-lkp@intel.com/ (local)

New smatch warnings:
net/xfrm/xfrm_user.c:3299 xfrm_do_migrate_state() warn: missing error code? 'err'
Looking at this more closely, xfrm_user_state_lookup() always sets *errp 
when it returns NULL.
Old smatch warnings:
net/xfrm/xfrm_user.c:1024 xfrm_add_sa() warn: missing error code? 'err'
net/xfrm/xfrm_user.c:2248 xfrm_add_policy() warn: missing error code? 'err'
net/xfrm/xfrm_user.c:3018 xfrm_add_acquire() warn: missing error code 'err'
Also, as the "Old smatch warnings" show, this same pattern exists elsewhere
in the file, most of the calls to xfrm_user_state_lookup().

I'm inclined to leave it as is rather than change a pattern that's 
consistent throughout the file. Does smatch follow the code doing
cross-function analysis? In this case, look into xfrm_user_state_lookup()
and further down to see that *errp is set when NULL is returned?

Thanks
-antony
vim +/err +3299 net/xfrm/xfrm_user.c

d3019c1db87425 Antony Antony 2026-01-17  3240  static int xfrm_do_migrate_state(struct sk_buff *skb, struct nlmsghdr *nlh,
d3019c1db87425 Antony Antony 2026-01-17  3241  				 struct nlattr **attrs, struct netlink_ext_ack *extack)
d3019c1db87425 Antony Antony 2026-01-17  3242  {
d3019c1db87425 Antony Antony 2026-01-17  3243  	int err = -ESRCH;
d3019c1db87425 Antony Antony 2026-01-17  3244  	struct xfrm_state *x;
d3019c1db87425 Antony Antony 2026-01-17  3245  	struct net *net = sock_net(skb->sk);
d3019c1db87425 Antony Antony 2026-01-17  3246  	struct xfrm_encap_tmpl *encap = NULL;
d3019c1db87425 Antony Antony 2026-01-17  3247  	struct xfrm_user_offload *xuo = NULL;
d3019c1db87425 Antony Antony 2026-01-17  3248  	struct xfrm_migrate m = { .old_saddr.a4 = 0,};
d3019c1db87425 Antony Antony 2026-01-17  3249  	struct xfrm_user_migrate_state *um = nlmsg_data(nlh);
d3019c1db87425 Antony Antony 2026-01-17  3250  
d3019c1db87425 Antony Antony 2026-01-17  3251  	if (!um->id.spi) {
d3019c1db87425 Antony Antony 2026-01-17  3252  		NL_SET_ERR_MSG(extack, "Invalid SPI 0x0");
d3019c1db87425 Antony Antony 2026-01-17  3253  		return -EINVAL;
d3019c1db87425 Antony Antony 2026-01-17  3254  	}
d3019c1db87425 Antony Antony 2026-01-17  3255  
d3019c1db87425 Antony Antony 2026-01-17  3256  	err = copy_from_user_migrate_state(&m, um);
d3019c1db87425 Antony Antony 2026-01-17  3257  	if (err)
d3019c1db87425 Antony Antony 2026-01-17  3258  		return err;
d3019c1db87425 Antony Antony 2026-01-17  3259  
d3019c1db87425 Antony Antony 2026-01-17  3260  	x = xfrm_user_state_lookup(net, &um->id, attrs, &err);
d3019c1db87425 Antony Antony 2026-01-17  3261  
d3019c1db87425 Antony Antony 2026-01-17  3262  	if (x) {
d3019c1db87425 Antony Antony 2026-01-17  3263  		struct xfrm_state *xc;
d3019c1db87425 Antony Antony 2026-01-17  3264  
d3019c1db87425 Antony Antony 2026-01-17  3265  		if (!x->dir) {
d3019c1db87425 Antony Antony 2026-01-17  3266  			NL_SET_ERR_MSG(extack, "State direction is invalid");
d3019c1db87425 Antony Antony 2026-01-17  3267  			err = -EINVAL;
d3019c1db87425 Antony Antony 2026-01-17  3268  			goto error;
d3019c1db87425 Antony Antony 2026-01-17  3269  		}
d3019c1db87425 Antony Antony 2026-01-17  3270  
d3019c1db87425 Antony Antony 2026-01-17  3271  		if (attrs[XFRMA_ENCAP]) {
d3019c1db87425 Antony Antony 2026-01-17  3272  			encap = kmemdup(nla_data(attrs[XFRMA_ENCAP]),
d3019c1db87425 Antony Antony 2026-01-17  3273  					sizeof(*encap), GFP_KERNEL);
d3019c1db87425 Antony Antony 2026-01-17  3274  			if (!encap) {
d3019c1db87425 Antony Antony 2026-01-17  3275  				err = -ENOMEM;
d3019c1db87425 Antony Antony 2026-01-17  3276  				goto error;
d3019c1db87425 Antony Antony 2026-01-17  3277  			}
d3019c1db87425 Antony Antony 2026-01-17  3278  		}
d3019c1db87425 Antony Antony 2026-01-17  3279  		if (attrs[XFRMA_OFFLOAD_DEV]) {
d3019c1db87425 Antony Antony 2026-01-17  3280  			xuo = kmemdup(nla_data(attrs[XFRMA_OFFLOAD_DEV]),
d3019c1db87425 Antony Antony 2026-01-17  3281  				      sizeof(*xuo), GFP_KERNEL);
d3019c1db87425 Antony Antony 2026-01-17  3282  			if (!xuo) {
d3019c1db87425 Antony Antony 2026-01-17  3283  				err = -ENOMEM;
d3019c1db87425 Antony Antony 2026-01-17  3284  				goto error;
d3019c1db87425 Antony Antony 2026-01-17  3285  			}
d3019c1db87425 Antony Antony 2026-01-17  3286  		}
d3019c1db87425 Antony Antony 2026-01-17  3287  		xc = xfrm_state_migrate(x, &m, encap, net, xuo, extack);
d3019c1db87425 Antony Antony 2026-01-17  3288  		if (xc) {
d3019c1db87425 Antony Antony 2026-01-17  3289  			xfrm_state_delete(x);
d3019c1db87425 Antony Antony 2026-01-17  3290  			xfrm_send_migrate_state(um, encap, xuo);
d3019c1db87425 Antony Antony 2026-01-17  3291  			err = 0;
d3019c1db87425 Antony Antony 2026-01-17  3292  		} else {
d3019c1db87425 Antony Antony 2026-01-17  3293  			if (extack && !extack->_msg)
d3019c1db87425 Antony Antony 2026-01-17  3294  				NL_SET_ERR_MSG(extack, "State migration clone failed");
d3019c1db87425 Antony Antony 2026-01-17  3295  			err = -EINVAL;
d3019c1db87425 Antony Antony 2026-01-17  3296  		}
d3019c1db87425 Antony Antony 2026-01-17  3297  	} else {
d3019c1db87425 Antony Antony 2026-01-17  3298  		NL_SET_ERR_MSG(extack, "Can not find state");
d3019c1db87425 Antony Antony 2026-01-17 @3299  		return err;

s/err/-ESRCH/.  err is zero/success here.

d3019c1db87425 Antony Antony 2026-01-17  3300  	}
d3019c1db87425 Antony Antony 2026-01-17  3301  error:
d3019c1db87425 Antony Antony 2026-01-17  3302  	xfrm_state_put(x);
d3019c1db87425 Antony Antony 2026-01-17  3303  	kfree(encap);
d3019c1db87425 Antony Antony 2026-01-17  3304  	kfree(xuo);
d3019c1db87425 Antony Antony 2026-01-17  3305  	return err;
d3019c1db87425 Antony Antony 2026-01-17  3306  }

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki

-- 
Devel mailing list -- devel@lists.linux-ipsec.org
To unsubscribe send an email to devel-leave@lists.linux-ipsec.org
  
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help