[bug report] lib: update LZ4 compressor module

[bug report] lib: update LZ4 compressor module

[Dan Carpenter]

Hello Sven Schmidt,

This is a semi-automatic email about new static checker warnings.

The patch e355e356f87b: "lib: update LZ4 compressor module" from Jan 
25, 2017, leads to the following Smatch complaint:

lib/lz4/lz4_compress.c:890 lz4_compress()
	 warn: variable dereferenced before check 'dst_len' (see line 880)

lib/lz4/lz4_compress.c
   879		size_t *dst_len, void *wrkmem) {
   880		*dst_len = LZ4_compress_default(src, dst, (int)src_len,
                 ^^^^^^^
Dereference.

   881			(int)((size_t)dst_len), wrkmem);
   882	
   883		/*
   884		 * Prior lz4_compress will return -1 in case of error
   885		 * and 0 on success
   886		 * while new LZ4_compress_fast/default
   887		 * returns 0 in case of error
   888		 * and the output length on success
   889		 */
   890		if (!dst_len)
                     ^^^^^^^
Checked too late.  Can this test just be deleted?

   891			return -1;
   892		else

regards,
dan carpenter

Re: [bug report] lib: update LZ4 compressor module

[Dan Carpenter]

Same thing for lz4hc_compress() as well.

    lib/lz4/lz4hc_compress.c:622 lz4hc_compress()
    warn: variable dereferenced before check 'dst_len' (see line 612)

regards,
dan carpenter

Re: [bug report] lib: update LZ4 compressor module

[Sven Schmidt (4sschmid)]

Am 31. Januar 2017 11:43:28 MEZ schrieb Dan Carpenter <dan.carpenter@oracle.com>:
>Same thing for lz4hc_compress() as well.
>
>    lib/lz4/lz4hc_compress.c:622 lz4hc_compress()
>    warn: variable dereferenced before check 'dst_len' (see line 612)
>
>regards,
>dan carpenter

Hi Dan, 

I did not receive any previous E-Mails, so this one is a bit out of context for me. I checked the kernel-janitors archive though and found the E-Mail you may reference to. 

I will fix it.

Is there a way to automatically receive any following messages?

Thanks,

Sven

Re: [bug report] lib: update LZ4 compressor module

[Dan Carpenter]

On Tue, Jan 31, 2017 at 01:09:23PM +0100, Sven Schmidt (4sschmid) wrote:
> Am 31. Januar 2017 11:43:28 MEZ schrieb Dan Carpenter <dan.carpenter@oracle.com>:
> >Same thing for lz4hc_compress() as well.
> >
> >    lib/lz4/lz4hc_compress.c:622 lz4hc_compress()
> >    warn: variable dereferenced before check 'dst_len' (see line 612)
> >
> >regards,
> >dan carpenter
> 
> Hi Dan, 
> 
> I did not receive any previous E-Mails, so this one is a bit out of context for me. I checked the kernel-janitors archive though and found the E-Mail you may reference to. 
> 
> I will fix it.
> 
> Is there a way to automatically receive any following messages?
> 

It's a Smatch warning.

regards,
dan carpenter