Fix inefficiences and rename nla_strlcpy
Subsystems (13)
v5 cover Fix inefficiences and rename nla_strlcpy
From: Francis Laniel <[email protected]>
Hi.
I hope you are all fine and the same for your relatives.
This patch set answers to first three issues listed in:
https://github.com/KSPP/linux/issues/110
To sum up, the patch contributions are the following:
1. the first patch fixes an inefficiency where some bytes in dst were written twice, one with 0 the other with src content.
2. The second one modifies nla_strlcpy to return the same value as strscpy, i.e. number of bytes written or -E2BIG if src was truncated.
It also modifies code that calls nla_strlcpy and checks for its return value.
3. The third renames nla_strlcpy to nla_strscpy.
Unfortunately, I did not find how to create struct nlattr objects so I tested my modifications on simple char* and with GDB using tc to get to tcf_proto_check_kind.
If you see any way to improve the code or have any remark, feel free to comment.
Best regards and take care of yourselves.
Francis Laniel (3):
Fix unefficient call to memset before memcpu in nla_strlcpy. Modify return value of nla_strlcpy to match that of strscpy.
treewide: rename nla_strlcpy to nla_strscpy.
drivers/infiniband/core/nldev.c | 10 +++--- drivers/net/can/vxcan.c | 4 +-- drivers/net/veth.c | 4 +-- include/linux/genl_magic_struct.h | 2 +- include/net/netlink.h | 4 +-- include/net/pkt_cls.h | 2 +- kernel/taskstats.c | 2 +- lib/nlattr.c | 42 ++++++++++++++-------- net/core/fib_rules.c | 4 +-- net/core/rtnetlink.c | 12 +++---- net/decnet/dn_dev.c | 2 +- net/ieee802154/nl-mac.c | 2 +- net/ipv4/devinet.c | 2 +- net/ipv4/fib_semantics.c | 2 +- net/ipv4/metrics.c | 2 +- net/netfilter/ipset/ip_set_hash_netiface.c | 4 +-- net/netfilter/nf_tables_api.c | 6 ++-- net/netfilter/nfnetlink_acct.c | 2 +- net/netfilter/nfnetlink_cthelper.c | 4 +-- net/netfilter/nft_ct.c | 2 +- net/netfilter/nft_log.c | 2 +- net/netlabel/netlabel_mgmt.c | 2 +- net/nfc/netlink.c | 2 +- net/sched/act_api.c | 2 +- net/sched/act_ipt.c | 2 +- net/sched/act_simple.c | 4 +-- net/sched/cls_api.c | 2 +- net/sched/sch_api.c | 2 +- net/tipc/netlink_compat.c | 2 +- 29 files changed, 73 insertions(+), 61 deletions(-)
Signature · 1 line
2.20.1
PATCH v5 1/3 Fix unefficient call to memset before memcpu in nla_strlcpy.
From: Francis Laniel <[email protected]>
Before this commit, nla_strlcpy first memseted dst to 0 then wrote src into it. This is inefficient because bytes whom number is less than src length are written twice.
This patch solves this issue by first writing src into dst then fill dst with 0's.
Note that, in the case where src length is higher than dst, only 0 is written. Otherwise there are as many 0's written to fill dst.
For example, if src is "foo\0" and dst is 5 bytes long, the result will be:
1. "fooGG" after memcpy (G means garbage).
2. "foo\0\0" after memset.
--- lib/nlattr.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
lib/nlattr.c +2 -1
@@ -731,8 +731,9 @@ size_t nla_strlcpy(char *dst, const struct nlattr *nla, size_t dstsize)if (dstsize >0 ) {size_t len = (srclen >= dstsize) ? dstsize -1 : srclen; -memset (dst,0 , dstsize);memcpy (dst, src, len); +/* Zero pad end of dst. */ +memset (dst + len,0 , dstsize - len); }return srclen;
Signature · 1 line
2.20.1
PATCH v5 2/3 Modify return value of nla_strlcpy to match that of strscpy.
In reply to [email protected]From: Francis Laniel <[email protected]>
nla_strlcpy now returns -E2BIG if src was truncated when written to dst. It also returns this error value if dstsize is 0 or higher than INT_MAX.
For example, if src is "foo\0" and dst is 3 bytes long, the result will be:
1. "foG" after memcpy (G means garbage).
2. "fo\0" after memset.
3. -E2BIG is returned because src was not completely written into dst.
The callers of nla_strlcpy were modified to take into account this modification.
--- include/net/netlink.h | 2 +- include/net/pkt_cls.h | 2 +- lib/nlattr.c | 39 +++++++++++++++++++++++++-------------- net/sched/act_api.c | 2 +- net/sched/cls_api.c | 2 +- net/sched/sch_api.c | 2 +- 6 files changed, 30 insertions(+), 19 deletions(-)
include/net/netlink.h +1 -1
@@ -506,7 +506,7 @@ int __nla_parse(struct nlattr **tb, int maxtype, const struct nlattr *head, struct netlink_ext_ack *extack);int nla_policy_len (const struct nla_policy *,int );struct nlattr *nla_find (const struct nlattr *head,int len,int attrtype); -size_t nla_strlcpy (char *dst,const struct nlattr *nla,size_t dstsize); +ssize_t nla_strlcpy (char *dst,const struct nlattr *nla,size_t dstsize);char *nla_strdup (const struct nlattr *nla,gfp_t flags);int nla_memcpy (void *dest,const struct nlattr *src,int count);int nla_memcmp (const struct nlattr *nla,const void *data,size_t size);
include/net/pkt_cls.h +1 -1
@@ -512,7 +512,7 @@ tcf_change_indev(struct net *net, struct nlattr *indev_tlv,char indev[IFNAMSIZ ];struct net_device *dev; -if (nla_strlcpy (indev, indev_tlv,IFNAMSIZ ) >=IFNAMSIZ ) { +if (nla_strlcpy (indev, indev_tlv,IFNAMSIZ ) <0 ) {NL_SET_ERR_MSG_ATTR (extack, indev_tlv,"Interface name too long" );return -EINVAL ;
lib/nlattr.c +25 -14
@@ -710,33 +710,44 @@ EXPORT_SYMBOL(nla_find);/** * nla_strlcpy - Copy string attribute payload into a sized buffer -* @dst: where to copy the string to -* @nla: attribute to copy the string from -* @dstsize: size of destination buffer +* @dst: Where to copy the string to. +* @nla: Attribute to copy the string from. +* @dstsize: Size of destination buffer. * * Copies at most dstsize - 1 bytes into the destination buffer. -* The result is always a valid NUL-terminated string. Unlike -* strlcpy the destination buffer is always padded out. +* Unlike strlcpy the destination buffer is always padded out. * -* Returns the length of the source buffer. +* Return: +* * srclen - Returns @nla length (not including the trailing %NUL). +* * -E2BIG - If @dstsize is 0 or greater than U16_MAX or @nla length greater +* than @dstsize. */ -size_t nla_strlcpy (char *dst,const struct nlattr *nla,size_t dstsize) +ssize_t nla_strlcpy (char *dst,const struct nlattr *nla,size_t dstsize) {size_t srclen =nla_len (nla);char *src =nla_data (nla); +ssize_t ret; +size_t len; + +if (dstsize ==0 ||WARN_ON_ONCE (dstsize >U16_MAX )) +return -E2BIG ;if (srclen >0 && src[srclen -1 ] =='\0' ) srclen--; -if (dstsize >0 ) { -size_t len = (srclen >= dstsize) ? dstsize -1 : srclen; - -memcpy (dst, src, len); -/* Zero pad end of dst. */ -memset (dst + len,0 , dstsize - len); +if (srclen >= dstsize) { + len = dstsize -1 ; + ret = -E2BIG ; + }else { + len = srclen; + ret = len; } -return srclen; +memcpy (dst, src, len); +/* Zero pad end of dst. */ +memset (dst + len,0 , dstsize - len); + +return ret; }EXPORT_SYMBOL (nla_strlcpy);
net/sched/act_api.c +1 -1
@@ -935,7 +935,7 @@ struct tc_action *tcf_action_init_1(struct net *net, struct tcf_proto *tp,NL_SET_ERR_MSG (extack,"TC action kind must be specified" ); goto err_out; } -if (nla_strlcpy (act_name, kind,IFNAMSIZ ) >=IFNAMSIZ ) { +if (nla_strlcpy (act_name, kind,IFNAMSIZ ) <0 ) {NL_SET_ERR_MSG (extack,"TC action name too long" ); goto err_out; }
net/sched/cls_api.c +1 -1
@@ -223,7 +223,7 @@ static inline u32 tcf_auto_prio(struct tcf_proto *tp)static bool tcf_proto_check_kind (struct nlattr *kind,char *name) {if (kind) -return nla_strlcpy (name, kind,IFNAMSIZ ) >=IFNAMSIZ ; +return nla_strlcpy (name, kind,IFNAMSIZ ) <0 ;memset (name,0 ,IFNAMSIZ );return false; }
net/sched/sch_api.c +1 -1
@@ -1170,7 +1170,7 @@ static struct Qdisc *qdisc_create(struct net_device *dev,#ifdef CONFIG_MODULES if (ops ==NULL && kind !=NULL ) {char name[IFNAMSIZ ]; -if (nla_strlcpy (name, kind,IFNAMSIZ ) <IFNAMSIZ ) { +if (nla_strlcpy (name, kind,IFNAMSIZ ) >=0 ) {/* We dropped the RTNL semaphore in order to * perform the module load. So, even if we * succeeded in loading the module we have to
Signature · 1 line
2.20.1
PATCH v5 3/3 treewide: rename nla_strlcpy to nla_strscpy.
In reply to [email protected]From: Francis Laniel <[email protected]>
Calls to nla_strlcpy are now replaced by calls to nla_strscpy which is the new name of this function.
--- drivers/infiniband/core/nldev.c | 10 +++++----- drivers/net/can/vxcan.c | 4 ++-- drivers/net/veth.c | 4 ++-- include/linux/genl_magic_struct.h | 2 +- include/net/netlink.h | 4 ++-- include/net/pkt_cls.h | 2 +- kernel/taskstats.c | 2 +- lib/nlattr.c | 6 +++--- net/core/fib_rules.c | 4 ++-- net/core/rtnetlink.c | 12 ++++++------ net/decnet/dn_dev.c | 2 +- net/ieee802154/nl-mac.c | 2 +- net/ipv4/devinet.c | 2 +- net/ipv4/fib_semantics.c | 2 +- net/ipv4/metrics.c | 2 +- net/netfilter/ipset/ip_set_hash_netiface.c | 4 ++-- net/netfilter/nf_tables_api.c | 6 +++--- net/netfilter/nfnetlink_acct.c | 2 +- net/netfilter/nfnetlink_cthelper.c | 4 ++-- net/netfilter/nft_ct.c | 2 +- net/netfilter/nft_log.c | 2 +- net/netlabel/netlabel_mgmt.c | 2 +- net/nfc/netlink.c | 2 +- net/sched/act_api.c | 2 +- net/sched/act_ipt.c | 2 +- net/sched/act_simple.c | 4 ++-- net/sched/cls_api.c | 2 +- net/sched/sch_api.c | 2 +- net/tipc/netlink_compat.c | 2 +- 29 files changed, 49 insertions(+), 49 deletions(-)
drivers/infiniband/core/nldev.c +5 -5
@@ -932,7 +932,7 @@ static int nldev_set_doit(struct sk_buff *skb, struct nlmsghdr *nlh,if (tb[RDMA_NLDEV_ATTR_DEV_NAME ]) {char name[IB_DEVICE_NAME_MAX ] = {}; -nla_strlcpy (name, tb[RDMA_NLDEV_ATTR_DEV_NAME ], +nla_strscpy (name, tb[RDMA_NLDEV_ATTR_DEV_NAME ],IB_DEVICE_NAME_MAX );if (strlen (name) ==0 ) { err = -EINVAL ;
@@ -1529,13 +1529,13 @@ static int nldev_newlink(struct sk_buff *skb, struct nlmsghdr *nlh, !tb[RDMA_NLDEV_ATTR_LINK_TYPE] || !tb[RDMA_NLDEV_ATTR_NDEV_NAME])return -EINVAL ; -nla_strlcpy (ibdev_name, tb[RDMA_NLDEV_ATTR_DEV_NAME ], +nla_strscpy (ibdev_name, tb[RDMA_NLDEV_ATTR_DEV_NAME ],sizeof (ibdev_name));if (strchr (ibdev_name,'%' ) ||strlen (ibdev_name) ==0 )return -EINVAL ; -nla_strlcpy (type, tb[RDMA_NLDEV_ATTR_LINK_TYPE ],sizeof (type)); -nla_strlcpy (ndev_name, tb[RDMA_NLDEV_ATTR_NDEV_NAME ], +nla_strscpy (type, tb[RDMA_NLDEV_ATTR_LINK_TYPE ],sizeof (type)); +nla_strscpy (ndev_name, tb[RDMA_NLDEV_ATTR_NDEV_NAME ],sizeof (ndev_name)); ndev =dev_get_by_name (sock_net (skb->sk), ndev_name);
@@ -1602,7 +1602,7 @@ static int nldev_get_chardev(struct sk_buff *skb, struct nlmsghdr *nlh,if (err || !tb[RDMA_NLDEV_ATTR_CHARDEV_TYPE ])return -EINVAL ; -nla_strlcpy (client_name, tb[RDMA_NLDEV_ATTR_CHARDEV_TYPE ], +nla_strscpy (client_name, tb[RDMA_NLDEV_ATTR_CHARDEV_TYPE ],sizeof (client_name));if (tb[RDMA_NLDEV_ATTR_DEV_INDEX ]) {
drivers/net/can/vxcan.c +2 -2
@@ -186,7 +186,7 @@ static int vxcan_newlink(struct net *net, struct net_device *dev, }if (ifmp && tbp[IFLA_IFNAME ]) { -nla_strlcpy (ifname, tbp[IFLA_IFNAME ],IFNAMSIZ ); +nla_strscpy (ifname, tbp[IFLA_IFNAME ],IFNAMSIZ ); name_assign_type =NET_NAME_USER ; }else {snprintf (ifname,IFNAMSIZ ,DRV_NAME "%%d" );
@@ -223,7 +223,7 @@ static int vxcan_newlink(struct net *net, struct net_device *dev,/* register first device */ if (tb[IFLA_IFNAME ]) -nla_strlcpy (dev->name, tb[IFLA_IFNAME ],IFNAMSIZ ); +nla_strscpy (dev->name, tb[IFLA_IFNAME ],IFNAMSIZ );else snprintf (dev->name,IFNAMSIZ ,DRV_NAME "%%d" );
drivers/net/veth.c +2 -2
@@ -1329,7 +1329,7 @@ static int veth_newlink(struct net *src_net, struct net_device *dev, }if (ifmp && tbp[IFLA_IFNAME ]) { -nla_strlcpy (ifname, tbp[IFLA_IFNAME ],IFNAMSIZ ); +nla_strscpy (ifname, tbp[IFLA_IFNAME ],IFNAMSIZ ); name_assign_type =NET_NAME_USER ; }else {snprintf (ifname,IFNAMSIZ ,DRV_NAME "%%d" );
@@ -1379,7 +1379,7 @@ static int veth_newlink(struct net *src_net, struct net_device *dev,eth_hw_addr_random (dev);if (tb[IFLA_IFNAME ]) -nla_strlcpy (dev->name, tb[IFLA_IFNAME ],IFNAMSIZ ); +nla_strscpy (dev->name, tb[IFLA_IFNAME ],IFNAMSIZ );else snprintf (dev->name,IFNAMSIZ ,DRV_NAME "%%d" );
include/linux/genl_magic_struct.h +1 -1
@@ -89,7 +89,7 @@ static inline int nla_put_u64_0pad(struct sk_buff *skb, int attrtype, u64 value) nla_get_u64, nla_put_u64_0pad, false)#define __str_field (attr_nr, attr_flag, name, maxlen) \ __array(attr_nr, attr_flag, name, NLA_NUL_STRING, char, maxlen, \ - nla_strlcpy, nla_put, false) + nla_strscpy, nla_put, false)#define __bin_field (attr_nr, attr_flag, name, maxlen) \ __array(attr_nr, attr_flag, name, NLA_BINARY, char, maxlen, \ nla_memcpy, nla_put, false)
include/net/netlink.h +2 -2
@@ -142,7 +142,7 @@* Attribute Misc: * nla_memcpy(dest, nla, count) copy attribute into memory * nla_memcmp(nla, data, size) compare attribute with memory area -* nla_strlcpy(dst, nla, size) copy attribute to a sized string +* nla_strscpy(dst, nla, size) copy attribute to a sized string * nla_strcmp(nla, str) compare attribute with string * * Attribute Parsing:
@@ -506,7 +506,7 @@ int __nla_parse(struct nlattr **tb, int maxtype, const struct nlattr *head, struct netlink_ext_ack *extack);int nla_policy_len (const struct nla_policy *,int );struct nlattr *nla_find (const struct nlattr *head,int len,int attrtype); -ssize_t nla_strlcpy (char *dst,const struct nlattr *nla,size_t dstsize); +ssize_t nla_strscpy (char *dst,const struct nlattr *nla,size_t dstsize);char *nla_strdup (const struct nlattr *nla,gfp_t flags);int nla_memcpy (void *dest,const struct nlattr *src,int count);int nla_memcmp (const struct nlattr *nla,const void *data,size_t size);
include/net/pkt_cls.h +1 -1
@@ -512,7 +512,7 @@ tcf_change_indev(struct net *net, struct nlattr *indev_tlv,char indev[IFNAMSIZ ];struct net_device *dev; -if (nla_strlcpy (indev, indev_tlv,IFNAMSIZ ) <0 ) { +if (nla_strscpy (indev, indev_tlv,IFNAMSIZ ) <0 ) {NL_SET_ERR_MSG_ATTR (extack, indev_tlv,"Interface name too long" );return -EINVAL ;
kernel/taskstats.c +1 -1
@@ -346,7 +346,7 @@ static int parse(struct nlattr *na, struct cpumask *mask) data =kmalloc (len,GFP_KERNEL );if (!data)return -ENOMEM ; -nla_strlcpy (data, na, len); +nla_strscpy (data, na, len); ret =cpulist_parse (data, mask);kfree (data);return ret;
lib/nlattr.c +3 -3
@@ -709,7 +709,7 @@ struct nlattr *nla_find(const struct nlattr *head, int len, int attrtype)EXPORT_SYMBOL (nla_find);/** -* nla_strlcpy - Copy string attribute payload into a sized buffer +* nla_strscpy - Copy string attribute payload into a sized buffer * @dst: Where to copy the string to. * @nla: Attribute to copy the string from. * @dstsize: Size of destination buffer.
@@ -722,7 +722,7 @@ EXPORT_SYMBOL(nla_find);* * -E2BIG - If @dstsize is 0 or greater than U16_MAX or @nla length greater * than @dstsize. */ -ssize_t nla_strlcpy (char *dst,const struct nlattr *nla,size_t dstsize) +ssize_t nla_strscpy (char *dst,const struct nlattr *nla,size_t dstsize) {size_t srclen =nla_len (nla);char *src =nla_data (nla);
@@ -749,7 +749,7 @@ ssize_t nla_strlcpy(char *dst, const struct nlattr *nla, size_t dstsize)return ret; } -EXPORT_SYMBOL (nla_strlcpy); +EXPORT_SYMBOL (nla_strscpy);/** * nla_strdup - Copy string attribute payload into a newly allocated buffer
net/core/fib_rules.c +2 -2
@@ -563,7 +563,7 @@ static int fib_nl2rule(struct sk_buff *skb, struct nlmsghdr *nlh,struct net_device *dev; nlrule->iifindex =-1 ; -nla_strlcpy (nlrule->iifname, tb[FRA_IIFNAME ],IFNAMSIZ ); +nla_strscpy (nlrule->iifname, tb[FRA_IIFNAME ],IFNAMSIZ ); dev =__dev_get_by_name (net, nlrule->iifname);if (dev) nlrule->iifindex = dev->ifindex;
@@ -573,7 +573,7 @@ static int fib_nl2rule(struct sk_buff *skb, struct nlmsghdr *nlh,struct net_device *dev; nlrule->oifindex =-1 ; -nla_strlcpy (nlrule->oifname, tb[FRA_OIFNAME ],IFNAMSIZ ); +nla_strscpy (nlrule->oifname, tb[FRA_OIFNAME ],IFNAMSIZ ); dev =__dev_get_by_name (net, nlrule->oifname);if (dev) nlrule->oifindex = dev->ifindex;
net/core/rtnetlink.c +6 -6
@@ -1939,7 +1939,7 @@ static const struct rtnl_link_ops *linkinfo_to_kind_ops(const struct nlattr *nlaif (linfo[IFLA_INFO_KIND ]) {char kind[MODULE_NAME_LEN ]; -nla_strlcpy (kind, linfo[IFLA_INFO_KIND ],sizeof (kind)); +nla_strscpy (kind, linfo[IFLA_INFO_KIND ],sizeof (kind)); ops =rtnl_link_ops_get (kind); }
@@ -2953,9 +2953,9 @@ static struct net_device *rtnl_dev_get(struct net *net,if (!ifname) { ifname = buffer;if (ifname_attr) -nla_strlcpy (ifname, ifname_attr,IFNAMSIZ ); +nla_strscpy (ifname, ifname_attr,IFNAMSIZ );else if (altifname_attr) -nla_strlcpy (ifname, altifname_attr,ALTIFNAMSIZ ); +nla_strscpy (ifname, altifname_attr,ALTIFNAMSIZ );else return NULL ; }
@@ -2983,7 +2983,7 @@ static int rtnl_setlink(struct sk_buff *skb, struct nlmsghdr *nlh, goto errout;if (tb[IFLA_IFNAME ]) -nla_strlcpy (ifname, tb[IFLA_IFNAME ],IFNAMSIZ ); +nla_strscpy (ifname, tb[IFLA_IFNAME ],IFNAMSIZ );else ifname[0 ] ='\0' ;
@@ -3264,7 +3264,7 @@ static int __rtnl_newlink(struct sk_buff *skb, struct nlmsghdr *nlh,return err;if (tb[IFLA_IFNAME ]) -nla_strlcpy (ifname, tb[IFLA_IFNAME ],IFNAMSIZ ); +nla_strscpy (ifname, tb[IFLA_IFNAME ],IFNAMSIZ );else ifname[0 ] ='\0' ;
@@ -3296,7 +3296,7 @@ static int __rtnl_newlink(struct sk_buff *skb, struct nlmsghdr *nlh,memset (linkinfo,0 ,sizeof (linkinfo));if (linkinfo[IFLA_INFO_KIND ]) { -nla_strlcpy (kind, linkinfo[IFLA_INFO_KIND ],sizeof (kind)); +nla_strscpy (kind, linkinfo[IFLA_INFO_KIND ],sizeof (kind)); ops =rtnl_link_ops_get (kind); }else { kind[0 ] ='\0' ;
net/decnet/dn_dev.c +1 -1
@@ -658,7 +658,7 @@ static int dn_nl_newaddr(struct sk_buff *skb, struct nlmsghdr *nlh, ifa->ifa_dev = dn_db;if (tb[IFA_LABEL ]) -nla_strlcpy (ifa->ifa_label, tb[IFA_LABEL ],IFNAMSIZ ); +nla_strscpy (ifa->ifa_label, tb[IFA_LABEL ],IFNAMSIZ );else memcpy (ifa->ifa_label, dev->name,IFNAMSIZ );
net/ieee802154/nl-mac.c +1 -1
@@ -149,7 +149,7 @@ static struct net_device *ieee802154_nl_get_dev(struct genl_info *info)if (info->attrs[IEEE802154_ATTR_DEV_NAME ]) {char name[IFNAMSIZ +1 ]; -nla_strlcpy (name, info->attrs[IEEE802154_ATTR_DEV_NAME ], +nla_strscpy (name, info->attrs[IEEE802154_ATTR_DEV_NAME ],sizeof (name)); dev =dev_get_by_name (&init_net, name); }else if (info->attrs[IEEE802154_ATTR_DEV_INDEX ]) {
net/ipv4/devinet.c +1 -1
@@ -881,7 +881,7 @@ static struct in_ifaddr *rtm_to_ifaddr(struct net *net, struct nlmsghdr *nlh, ifa->ifa_broadcast =nla_get_in_addr (tb[IFA_BROADCAST ]);if (tb[IFA_LABEL ]) -nla_strlcpy (ifa->ifa_label, tb[IFA_LABEL ],IFNAMSIZ ); +nla_strscpy (ifa->ifa_label, tb[IFA_LABEL ],IFNAMSIZ );else memcpy (ifa->ifa_label, dev->name,IFNAMSIZ );
net/ipv4/fib_semantics.c +1 -1
@@ -973,7 +973,7 @@ bool fib_metrics_match(struct fib_config *cfg, struct fib_info *fi)char tmp[TCP_CA_NAME_MAX ];bool ecn_ca = false; -nla_strlcpy (tmp, nla,sizeof (tmp)); +nla_strscpy (tmp, nla,sizeof (tmp)); val =tcp_ca_get_key_by_name (fi->fib_net, tmp, &ecn_ca); }else {if (nla_len (nla) !=sizeof (u32))
net/ipv4/metrics.c +1 -1
@@ -31,7 +31,7 @@ static int ip_metrics_convert(struct net *net, struct nlattr *fc_mx,if (type ==RTAX_CC_ALGO ) {char tmp[TCP_CA_NAME_MAX ]; -nla_strlcpy (tmp, nla,sizeof (tmp)); +nla_strscpy (tmp, nla,sizeof (tmp)); val =tcp_ca_get_key_by_name (net, tmp, &ecn_ca);if (val ==TCP_CA_UNSPEC ) {NL_SET_ERR_MSG (extack,"Unknown tcp congestion algorithm" );
net/netfilter/ipset/ip_set_hash_netiface.c +2 -2
@@ -225,7 +225,7 @@ hash_netiface4_uadt(struct ip_set *set, struct nlattr *tb[],if (e.cidr >HOST_MASK )return -IPSET_ERR_INVALID_CIDR ; } -nla_strlcpy (e.iface, tb[IPSET_ATTR_IFACE ],IFNAMSIZ ); +nla_strscpy (e.iface, tb[IPSET_ATTR_IFACE ],IFNAMSIZ );if (tb[IPSET_ATTR_CADT_FLAGS ]) {u32 cadt_flags =ip_set_get_h32 (tb[IPSET_ATTR_CADT_FLAGS ]);
@@ -442,7 +442,7 @@ hash_netiface6_uadt(struct ip_set *set, struct nlattr *tb[],ip6_netmask (&e.ip, e.cidr); -nla_strlcpy (e.iface, tb[IPSET_ATTR_IFACE ],IFNAMSIZ ); +nla_strscpy (e.iface, tb[IPSET_ATTR_IFACE ],IFNAMSIZ );if (tb[IPSET_ATTR_CADT_FLAGS ]) {u32 cadt_flags =ip_set_get_h32 (tb[IPSET_ATTR_CADT_FLAGS ]);
net/netfilter/nf_tables_api.c +3 -3
@@ -1281,7 +1281,7 @@ static struct nft_chain *nft_chain_lookup(struct net *net,if (nla ==NULL )return ERR_PTR (-EINVAL ); -nla_strlcpy (search, nla,sizeof (search)); +nla_strscpy (search, nla,sizeof (search));WARN_ON (!rcu_read_lock_held () && !lockdep_commit_lock_is_held (net));
@@ -1721,7 +1721,7 @@ static struct nft_hook *nft_netdev_hook_alloc(struct net *net, goto err_hook_alloc; } -nla_strlcpy (ifname, attr,IFNAMSIZ ); +nla_strscpy (ifname, attr,IFNAMSIZ ); dev =__dev_get_by_name (net, ifname);if (!dev) { err = -ENOENT ;
@@ -5734,7 +5734,7 @@ struct nft_object *nft_obj_lookup(const struct net *net,struct rhlist_head *tmp, *list;struct nft_object *obj; -nla_strlcpy (search, nla,sizeof (search)); +nla_strscpy (search, nla,sizeof (search)); k.name = search; WARN_ON_ONCE(!rcu_read_lock_held() &&
net/netfilter/nfnetlink_acct.c +1 -1
@@ -112,7 +112,7 @@ static int nfnl_acct_new(struct net *net, struct sock *nfnl, nfacct->flags = flags; } -nla_strlcpy (nfacct->name, tb[NFACCT_NAME ],NFACCT_NAME_MAX ); +nla_strscpy (nfacct->name, tb[NFACCT_NAME ],NFACCT_NAME_MAX );if (tb[NFACCT_BYTES ]) { atomic64_set(&nfacct->bytes,
net/netfilter/nfnetlink_cthelper.c +2 -2
@@ -146,7 +146,7 @@ nfnl_cthelper_expect_policy(struct nf_conntrack_expect_policy *expect_policy, !tb[NFCTH_POLICY_EXPECT_TIMEOUT])return -EINVAL ; -nla_strlcpy (expect_policy->name, +nla_strscpy (expect_policy->name, tb[NFCTH_POLICY_NAME ],NF_CT_HELPER_NAME_LEN ); expect_policy->max_expected =ntohl (nla_get_be32 (tb[NFCTH_POLICY_EXPECT_MAX ]));
@@ -233,7 +233,7 @@ nfnl_cthelper_create(const struct nlattr * const tb[],if (ret <0 ) goto err1; -nla_strlcpy (helper->name, +nla_strscpy (helper->name, tb[NFCTH_NAME ],NF_CT_HELPER_NAME_LEN ); size =ntohl (nla_get_be32 (tb[NFCTH_PRIV_DATA_LEN ])); if (size > sizeof_field(struct nf_conn_help, data)) {
net/netfilter/nft_ct.c +1 -1
@@ -990,7 +990,7 @@ static int nft_ct_helper_obj_init(const struct nft_ctx *ctx,if (!priv->l4proto)return -ENOENT ; -nla_strlcpy (name, tb[NFTA_CT_HELPER_NAME ],sizeof (name)); +nla_strscpy (name, tb[NFTA_CT_HELPER_NAME ],sizeof (name));if (tb[NFTA_CT_HELPER_L3PROTO ]) family =ntohs (nla_get_be16 (tb[NFTA_CT_HELPER_L3PROTO ]));
net/netfilter/nft_log.c +1 -1
@@ -152,7 +152,7 @@ static int nft_log_init(const struct nft_ctx *ctx, priv->prefix =kmalloc (nla_len (nla) +1 ,GFP_KERNEL );if (priv->prefix ==NULL )return -ENOMEM ; -nla_strlcpy (priv->prefix, nla,nla_len (nla) +1 ); +nla_strscpy (priv->prefix, nla,nla_len (nla) +1 ); }else { priv->prefix = (char *)nft_log_null_prefix; }
net/netlabel/netlabel_mgmt.c +1 -1
@@ -95,7 +95,7 @@ static int netlbl_mgmt_add_common(struct genl_info *info, ret_val = -ENOMEM ; goto add_free_entry; } -nla_strlcpy (entry->domain, +nla_strscpy (entry->domain, info->attrs[NLBL_MGMT_A_DOMAIN ], tmp_size); }
net/nfc/netlink.c +1 -1
@@ -1226,7 +1226,7 @@ static int nfc_genl_fw_download(struct sk_buff *skb, struct genl_info *info)if (!dev)return -ENODEV ; -nla_strlcpy (firmware_name, info->attrs[NFC_ATTR_FIRMWARE_NAME ], +nla_strscpy (firmware_name, info->attrs[NFC_ATTR_FIRMWARE_NAME ],sizeof (firmware_name)); rc =nfc_fw_download (dev, firmware_name);
net/sched/act_api.c +1 -1
@@ -935,7 +935,7 @@ struct tc_action *tcf_action_init_1(struct net *net, struct tcf_proto *tp,NL_SET_ERR_MSG (extack,"TC action kind must be specified" ); goto err_out; } -if (nla_strlcpy (act_name, kind,IFNAMSIZ ) <0 ) { +if (nla_strscpy (act_name, kind,IFNAMSIZ ) <0 ) {NL_SET_ERR_MSG (extack,"TC action name too long" ); goto err_out; }
net/sched/act_ipt.c +1 -1
@@ -166,7 +166,7 @@ static int __tcf_ipt_init(struct net *net, unsigned int id, struct nlattr *nla,if (unlikely (!tname)) goto err1;if (tb[TCA_IPT_TABLE ] ==NULL || -nla_strlcpy (tname, tb[TCA_IPT_TABLE ],IFNAMSIZ ) >=IFNAMSIZ ) +nla_strscpy (tname, tb[TCA_IPT_TABLE ],IFNAMSIZ ) >=IFNAMSIZ )strcpy (tname,"mangle" ); t =kmemdup (td, td->u.target_size,GFP_KERNEL );
net/sched/act_simple.c +2 -2
@@ -52,7 +52,7 @@ static int alloc_defdata(struct tcf_defact *d, const struct nlattr *defdata) d->tcfd_defdata =kzalloc (SIMP_MAX_DATA ,GFP_KERNEL );if (unlikely (!d->tcfd_defdata))return -ENOMEM ; -nla_strlcpy (d->tcfd_defdata, defdata,SIMP_MAX_DATA ); +nla_strscpy (d->tcfd_defdata, defdata,SIMP_MAX_DATA );return 0 ; }
@@ -71,7 +71,7 @@ static int reset_policy(struct tc_action *a, const struct nlattr *defdata,spin_lock_bh (&d->tcf_lock); goto_ch =tcf_action_set_ctrlact (a, p->action, goto_ch);memset (d->tcfd_defdata,0 ,SIMP_MAX_DATA ); -nla_strlcpy (d->tcfd_defdata, defdata,SIMP_MAX_DATA ); +nla_strscpy (d->tcfd_defdata, defdata,SIMP_MAX_DATA );spin_unlock_bh (&d->tcf_lock);if (goto_ch)tcf_chain_put_by_act (goto_ch);
net/sched/cls_api.c +1 -1
@@ -223,7 +223,7 @@ static inline u32 tcf_auto_prio(struct tcf_proto *tp)static bool tcf_proto_check_kind (struct nlattr *kind,char *name) {if (kind) -return nla_strlcpy (name, kind,IFNAMSIZ ) <0 ; +return nla_strscpy (name, kind,IFNAMSIZ ) <0 ;memset (name,0 ,IFNAMSIZ );return false; }
net/sched/sch_api.c +1 -1
@@ -1170,7 +1170,7 @@ static struct Qdisc *qdisc_create(struct net_device *dev,#ifdef CONFIG_MODULES if (ops ==NULL && kind !=NULL ) {char name[IFNAMSIZ ]; -if (nla_strlcpy (name, kind,IFNAMSIZ ) >=0 ) { +if (nla_strscpy (name, kind,IFNAMSIZ ) >=0 ) {/* We dropped the RTNL semaphore in order to * perform the module load. So, even if we * succeeded in loading the module we have to
net/tipc/netlink_compat.c +1 -1
@@ -695,7 +695,7 @@ static int tipc_nl_compat_link_dump(struct tipc_nl_compat_msg *msg, link_info.dest =nla_get_flag (link[TIPC_NLA_LINK_DEST ]); link_info.up =htonl (nla_get_flag (link[TIPC_NLA_LINK_UP ])); -nla_strlcpy (link_info.str, link[TIPC_NLA_LINK_NAME ], +nla_strscpy (link_info.str, link[TIPC_NLA_LINK_NAME ],TIPC_MAX_LINK_NAME ); return tipc_add_tlv(msg->rep, TIPC_TLV_LINK_INFO,
Signature · 1 line
2.20.1
Quoted 25 lines · On Fri, Nov 13, 2020 at 12:11:30PM +0100, [email protected]…
From: Francis Laniel <[email protected]>
Hi.
I hope you are all fine and the same for your relatives.
This patch set answers to first three issues listed in:
https://github.com/KSPP/linux/issues/110To sum up, the patch contributions are the following:
1. the first patch fixes an inefficiency where some bytes in dst were written twice, one with 0 the other with src content.
2. The second one modifiesnla_strlcpyto return the same value as strscpy, i.e. number of bytes written or-E2BIGif src was truncated.
It also modifies code that callsnla_strlcpyand checks for its return value.
3. The third renamesnla_strlcpytonla_strscpy.Unfortunately, I did not find how to create struct nlattr objects so I tested my modifications on simple char* and with GDB using tc to get to
tcf_proto_check_kind.If you see any way to improve the code or have any remark, feel free to comment.
Quoted 6 lines
Best regards and take care of yourselves.
Francis Laniel (3):
Fix unefficient call to memset before memcpu innla_strlcpy. Modify return value ofnla_strlcpyto match that of strscpy.
treewide: renamenla_strlcpytonla_strscpy.
Thanks! This looks good to me.
Jakub, does this look ready to you?
Quoted 35 lines
drivers/infiniband/core/nldev.c | 10 +++--- drivers/net/can/vxcan.c | 4 +-- drivers/net/veth.c | 4 +-- include/linux/genl_magic_struct.h | 2 +- include/net/netlink.h | 4 +-- include/net/pkt_cls.h | 2 +- kernel/taskstats.c | 2 +- lib/nlattr.c | 42 ++++++++++++++-------- net/core/fib_rules.c | 4 +-- net/core/rtnetlink.c | 12 +++---- net/decnet/dn_dev.c | 2 +- net/ieee802154/nl-mac.c | 2 +- net/ipv4/devinet.c | 2 +- net/ipv4/fib_semantics.c | 2 +- net/ipv4/metrics.c | 2 +- net/netfilter/ipset/ip_set_hash_netiface.c | 4 +-- net/netfilter/nf_tables_api.c | 6 ++-- net/netfilter/nfnetlink_acct.c | 2 +- net/netfilter/nfnetlink_cthelper.c | 4 +-- net/netfilter/nft_ct.c | 2 +- net/netfilter/nft_log.c | 2 +- net/netlabel/netlabel_mgmt.c | 2 +- net/nfc/netlink.c | 2 +- net/sched/act_api.c | 2 +- net/sched/act_ipt.c | 2 +- net/sched/act_simple.c | 4 +-- net/sched/cls_api.c | 2 +- net/sched/sch_api.c | 2 +- net/tipc/netlink_compat.c | 2 +- 29 files changed, 73 insertions(+), 61 deletions(-)Signature · 1 line
2.20.1
Signature · 1 line
Kees Cook
Jakub Kicinski
maintainerQuoted 3 lines · On Fri, 13 Nov 2020 10:56:26 -0800 Kees Cook wrote:
Thanks! This looks good to me.
Jakub, does this look ready to you?
Yup, looks good, sorry!
But it didn't get into patchwork cleanly :/
One more resend please? (assuming we're expected to take this into net-next)
Quoted 11 lines · Le dimanche 15 novembre 2020, 01:18:37 CET Jakub Kicinski a écrit :
Quoted 3 lines · On Fri, 13 Nov 2020 10:56:26 -0800 Kees Cook wrote:
Thanks! This looks good to me.
Jakub, does this look ready to you?
Yup, looks good, sorry!
But it didn't get into patchwork cleanly :/
One more resend please? (assuming we're expected to take this into net-next)
I will send it again and tag it for net-next.
Just to know, is patchwork this:
https://patchwork.kernel.org/project/netdevbpf/list/
?
Jakub Kicinski
maintainerQuoted 10 lines · On Sun, 15 Nov 2020 18:05:39 +0100 Francis Laniel wrote:
Quoted 9 lines · Le dimanche 15 novembre 2020, 01:18:37 CET Jakub Kicinski a écrit :
Quoted 3 lines · On Fri, 13 Nov 2020 10:56:26 -0800 Kees Cook wrote:
Thanks! This looks good to me.
Jakub, does this look ready to you?
Yup, looks good, sorry!
But it didn't get into patchwork cleanly :/
Quoted 7 lines
Quoted 2 lines
One more resend please? (assuming we're expected to take this into net-next)
I will send it again and tag it for net-next.
Just to know, is patchwork this:
https://patchwork.kernel.org/project/netdevbpf/list/
Yes, that's the one.
Quoted 14 lines · Le lundi 16 novembre 2020, 17:06:52 CET Jakub Kicinski a écrit :
Quoted 13 lines · On Sun, 15 Nov 2020 18:05:39 +0100 Francis Laniel wrote:
Quoted 11 lines · Le dimanche 15 novembre 2020, 01:18:37 CET Jakub Kicinski a écrit :
Quoted 3 lines · On Fri, 13 Nov 2020 10:56:26 -0800 Kees Cook wrote:
Thanks! This looks good to me.
Jakub, does this look ready to you?
Yup, looks good, sorry!
But it didn't get into patchwork cleanly :/
One more resend please? (assuming we're expected to take this into net-next)
Quoted 6 lines
Quoted 4 lines
I will send it again and tag it for net-next.
Just to know, is patchwork this:
https://patchwork.kernel.org/project/netdevbpf/list/Yes, that's the one.
OK! The tool seems cool.
So normally I resend the patches, but I do not see them in patchwork. Is there something am I supposed to do so they can be visible through patchwork?
Jakub Kicinski
maintainerQuoted 20 lines · On Mon, 16 Nov 2020 18:06:33 +0100 Francis Laniel wrote:
Quoted 19 lines · Le lundi 16 novembre 2020, 17:06:52 CET Jakub Kicinski a écrit :
Quoted 17 lines · On Sun, 15 Nov 2020 18:05:39 +0100 Francis Laniel wrote:
Quoted 11 lines · Le dimanche 15 novembre 2020, 01:18:37 CET Jakub Kicinski a écrit :
Quoted 3 lines · On Fri, 13 Nov 2020 10:56:26 -0800 Kees Cook wrote:
Thanks! This looks good to me.
Jakub, does this look ready to you?
Yup, looks good, sorry!
But it didn't get into patchwork cleanly :/
One more resend please? (assuming we're expected to take this into net-next)
I will send it again and tag it for net-next.
Just to know, is patchwork this:
https://patchwork.kernel.org/project/netdevbpf/list/
Quoted 7 lines
Quoted 1 line
Yes, that's the one.
OK! The tool seems cool.
So normally I resend the patches, but I do not see them in patchwork. Is there something am I supposed to do so they can be visible through patchwork?
They were visible, this is your resend:
https://patchwork.kernel.org/project/netdevbpf/list/?series=384457&state=*
(note the * at the end, sometimes it gets stripped when clicking on a link)
Maybe you checked after I already applied them, patches which had been "handled" (aka "Action required=no") are not displayed by default.
Quoted 31 lines · Le lundi 16 novembre 2020, 18:32:43 CET Jakub Kicinski a écrit :
Quoted 20 lines · On Mon, 16 Nov 2020 18:06:33 +0100 Francis Laniel wrote:
Quoted 19 lines · Le lundi 16 novembre 2020, 17:06:52 CET Jakub Kicinski a écrit :
Quoted 17 lines · On Sun, 15 Nov 2020 18:05:39 +0100 Francis Laniel wrote:
Quoted 11 lines · Le dimanche 15 novembre 2020, 01:18:37 CET Jakub Kicinski a écrit :
Quoted 3 lines · On Fri, 13 Nov 2020 10:56:26 -0800 Kees Cook wrote:
Thanks! This looks good to me.
Jakub, does this look ready to you?
Yup, looks good, sorry!
But it didn't get into patchwork cleanly :/
One more resend please? (assuming we're expected to take this into net-next)
I will send it again and tag it for net-next.
Just to know, is patchwork this:
https://patchwork.kernel.org/project/netdevbpf/list/Quoted 7 lines
Quoted 1 line
Yes, that's the one.
OK! The tool seems cool.
So normally I resend the patches, but I do not see them in patchwork. Is there something am I supposed to do so they can be visible through patchwork?
They were visible, this is your resend:
Quoted 7 lines
https://patchwork.kernel.org/project/netdevbpf/list/?series=384457&state=*
(note the * at the end, sometimes it gets stripped when clicking on a link)
Maybe you checked after I already applied them, patches which had been "handled" (aka "Action required=no") are not displayed by default.
Working of patchwork is now clearer for me, thank you! But yes I think I checked after you applied them.