Re: [ECOS] JFFS2 Issue.

[Jonathan Larmour <jifl at jifvik dot org>]

Ramgopal Kota wrote:
>  Hi,
>
> I am integrating JFFS2 into our MIPS Malta board. I am seeing a strange issue.
> In JFFS2 erase.c file and function jffs2_mark_erased_block , code section as below 
>
> struct jffs2_unknown_node marker = {
>       .magic =  cpu_to_je16(JFFS2_MAGIC_BITMASK),
>       .nodetype = cpu_to_je16(JFFS2_NODETYPE_CLEANMARKER), 
>       .totlen = cpu_to_je32(c->cleanmarker_size)
>     };
>
> struct jffs2_unknown_node
> {
>   /* All start like this */
>   jint16_t magic;
>   jint16_t nodetype;
>   jint32_t totlen; /* So we can skip over nodes we don't grok */
>   jint32_t hdr_crc;
> } __attribute__((packed));
>
> I am seeing that both .magic and .nodetype are filled with 0x2003 which is the value of JFFS2_NODETYPE_CLEANMARKER.
>
> If I change the code to the following it is working fine...
>
> struct jffs2_unknown_node marker = {
>       .magic =  cpu_to_je16(JFFS2_MAGIC_BITMASK),
>       .nodetype = cpu_to_je16(JFFS2_NODETYPE_CLEANMARKER), 
>       .totlen = cpu_to_je32(c->cleanmarker_size)
>     };
>
> marker.magic =  cpu_to_je16(JFFS2_MAGIC_BITMASK);
>
> I don't know if it is CFLAGS issue or Compiler issue.I am thinking it is a compiler optimisation re-ordering issue.I am using -Os flag.
> Compiler I am using is gcc version 3.2.1 (eCosCentric)

I recommend you eCos 3.0 and the updated GNU tools instead (you can't 
really use the updated tools with eCos much before 3.0).

You should have got an error compiling JFFS2 due to this in src/fs-ecos.c:

#if (__GNUC__ == 3) && (__GNUC_MINOR__ == 2) && \
    (defined (__arm__) || defined (_mips))
#error This compiler is known to be broken. Please see:
#error "http://ecos.sourceware.org/ml/ecos-patches/2003-08/msg00006.html";
#endif


Jifl
--
--["No sense being pessimistic, it wouldn't work anyway"]-- Opinions==mine

--
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss