linux-kernel - Re: [PATCH v2] libbpf: fix the name of a reused map

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <OSZP286MB1725F414B934C820252578CEB81C9@OSZP286MB1725.JPNP286.PROD.OUTLOOK.COM>
Date:   Wed, 13 Jul 2022 13:44:48 +0800
From:   Anquan Wu <leiqi96@...mail.com>
To:     Jiri Olsa <olsajiri@...il.com>
Cc:     andrii@...nel.org, daniel@...earbox.net, martin.lau@...ux.dev,
        song@...nel.org, yhs@...com, kpsingh@...nel.org,
        bpf@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2] libbpf: fix the name of a reused map

On Tue, 2022-07-12 at 10:31 +0200, Jiri Olsa wrote:
> On Tue, Jul 12, 2022 at 11:15:40AM +0800, Anquan Wu wrote:
> > BPF map name is limited to BPF_OBJ_NAME_LEN.
> > A map name is defined as being longer than BPF_OBJ_NAME_LEN,
> > it will be truncated to BPF_OBJ_NAME_LEN when a userspace program
> > calls libbpf to create the map. A pinned map also generates a path
> > in the /sys. If the previous program wanted to reuse the map，
> > it can not get bpf_map by name, because the name of the map is only
> > partially the same as the name which get from pinned path.
> > 
> > The syscall information below show that map name "process_pinned_map"
> > is truncated to "process_pinned_".
> > 
> >     bpf(BPF_OBJ_GET, {pathname="/sys/fs/bpf/process_pinned_map",
> >     bpf_fd=0, file_flags=0}, 144) = -1 ENOENT (No such file or
> > directory)
> > 
> >     bpf(BPF_MAP_CREATE, {map_type=BPF_MAP_TYPE_HASH, key_size=4,
> >     value_size=4,max_entries=1024, map_flags=0, inner_map_fd=0,
> >     map_name="process_pinned_",map_ifindex=0, btf_fd=3,
> > btf_key_type_id=6,
> >     btf_value_type_id=10,btf_vmlinux_value_type_id=0}, 72) = 4
> > 
> > This patch check that if the name of pinned map are the same as the
> > actual name for the first (BPF_OBJ_NAME_LEN - 1),
> > bpf map still uses the name which is included in bpf object.
> > 
> > Signed-off-by: Anquan Wu <leiqi96@...mail.com>
> > ---
> > 
> > v2: compare against zero explicitly
> > 
> > v1:
> > https://lore.kernel.org/linux-kernel/OSZP286MB1725A2361FA2EE8432C4D5F4B8879@OSZP286MB1725.JPNP286.PROD.OUTLOOK.COM/
> > ---
> >  tools/lib/bpf/libbpf.c | 8 +++++++-
> >  1 file changed, 7 insertions(+), 1 deletion(-)
> > 
> > diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c
> > index e89cc9c885b3..7b4d3604dfb4 100644
> > --- a/tools/lib/bpf/libbpf.c
> > +++ b/tools/lib/bpf/libbpf.c
> > @@ -4328,6 +4328,7 @@ int bpf_map__reuse_fd(struct bpf_map *map, int
> > fd)
> >  {
> >         struct bpf_map_info info = {};
> >         __u32 len = sizeof(info);
> > +       __u32 name_len;
> >         int new_fd, err;
> >         char *new_name;
> >  
> > @@ -4337,7 +4338,12 @@ int bpf_map__reuse_fd(struct bpf_map *map, int
> > fd)
> >         if (err)
> >                 return libbpf_err(err);
> >  
> > -       new_name = strdup(info.name);
> > +       name_len = strlen(info.name);
> > +       if (name_len == BPF_OBJ_NAME_LEN - 1 && strncmp(map->name,
> > info.name, name_len) == 0)
> 
> so what if the map->name is different after 'name_len' ?
> 
> jirka
> 

If  A map name is defined as being longer than name_len (name_len is 
"BPF_OBJ_NAME_LEN - 1" in this context), a program will fail to get a
reused bpf_map by bpf_object__find_map_by_name().
   
   fromhttps://github.com/libbpf/libbpf/blob/master/src/libbpf.c#L9295,
   pos->name in bpf_object__find_map_by_name() is from  new_name in      
   bpf_map_reuse_fd(). It can not find map by the name which is defined
   in bpf object.

I wrote some code to verify this problem and test the solution
mentioned above.
Link: https://github.com/leiqi96/libbpf-fix

Anquan


> > +               new_name = strdup(map->name);
> > +       else
> > +               new_name = strdup(info.name);
> > +
> >         if (!new_name)
> >                 return libbpf_err(-errno);
> >  
> > -- 
> > 2.32.0
> >