[REGRESSION] [PATCH] common: add blkcache init

[Angelo Dureghello]

Hi Marcel,

thanks for reporting,
the fix should be already in master, please let me know if it works on
your hardware.

Regards,
Angleo

On Mon, Jan 27, 2020 at 8:38 AM Marcel Ziswiler
<marcel.ziswiler at toradex.com> wrote:
>
> Hi Angelo
>
> On Tue, 2020-01-21 at 10:37 +0100, Angelo Dureghello wrote:
> > From: Angelo Durgehello <angelo.dureghello at timesys.com>
> >
> > On m68k, block_cache list is relocated, but next and prev list
> > pointers are not adjusted to the relocated struct list_head address,
> > so the first iteration over the block_cache list hangs.
> >
> > This patch initializes the block_cache list after relocation.
> >
> > Signed-off-by: Angelo Durgehello <angelo.dureghello at timesys.com>
> > Reviewed-by: Eric Nelson <eric at nelint.com>
> > ---
> > Changes for v2:
> > - call blkcache_init directly
> > ---
> >  common/board_r.c         | 3 +++
> >  drivers/block/blkcache.c | 9 ++++++++-
> >  include/blk.h            | 6 ++++++
> >  3 files changed, 17 insertions(+), 1 deletion(-)
> >
> > diff --git a/common/board_r.c b/common/board_r.c
> > index 8a0c1114e7..4f56c19fcc 100644
> > --- a/common/board_r.c
> > +++ b/common/board_r.c
> > @@ -864,6 +864,9 @@ static init_fnc_t init_sequence_r[] = {
> >  #endif
> >  #if defined(CONFIG_PRAM)
> >       initr_mem,
> > +#endif
> > +#ifdef CONFIG_BLOCK_CACHE
> > +     blkcache_init,
> >  #endif
> >       run_main_loop,
> >  };
> > diff --git a/drivers/block/blkcache.c b/drivers/block/blkcache.c
> > index 1fa64989d3..f603aa129d 100644
> > --- a/drivers/block/blkcache.c
> > +++ b/drivers/block/blkcache.c
> > @@ -21,13 +21,20 @@ struct block_cache_node {
> >       char *cache;
> >  };
> >
> > -static LIST_HEAD(block_cache);
> > +static struct list_head block_cache;
> >
> >  static struct block_cache_stats _stats = {
> >       .max_blocks_per_entry = 8,
> >       .max_entries = 32
> >  };
> >
> > +int blkcache_init(void)
> > +{
> > +     INIT_LIST_HEAD(&block_cache);
> > +
> > +     return 0;
> > +}
> > +
> >  static struct block_cache_node *cache_find(int iftype, int devnum,
> >                                          lbaint_t start, lbaint_t
> > blkcnt,
> >                                          unsigned long blksz)
> > diff --git a/include/blk.h b/include/blk.h
> > index d0c033aece..65db69f5d9 100644
> > --- a/include/blk.h
> > +++ b/include/blk.h
> > @@ -113,6 +113,12 @@ struct blk_desc {
> >       (PAD_SIZE(size, blk_desc->blksz))
> >
> >  #if CONFIG_IS_ENABLED(BLOCK_CACHE)
> > +
> > +/**
> > + * blkcache_init() - initialize the block cache list pointers
> > + */
> > +int blkcache_init(void);
> > +
> >  /**
> >   * blkcache_read() - attempt to read a set of blocks from cache
> >   *
>
> Unfortunately this one seems to break i.MX 8M Mini booting:
>
> U-Boot SPL 2020.01-00179-g35ac4f216b (Jan 25 2020 - 23:36:35 +0100)
> Normal Boot
> Trying to boot from MMC1
> NOTICE:  Configuring TZASC380
> NOTICE:  RDC off
> NOTICE:  BL31: v2.0(release):rel_imx_4.14.98_2.3.0-0-g09c5cc994-dirty
> NOTICE:  BL31: Built : 01:11:41, Jan 25 2020
> NOTICE:  sip svc init
>
>
> U-Boot 2020.01-00179-g35ac4f216b (Jan 25 2020 - 23:36:35 +0100)
>
> CPU:   Freescale i.MX8MMQ rev1.0 at 0 MHz
> Reset cause: POR
> DRAM:  2 GiB
> MMC:   FSL_SDHC: 0, FSL_SDHC: 1
> Loading Environment from MMC... "Synchronous Abort" handler, esr
> 0x96000004
> elr: 0000000040226de8 lr : 000000004021e99c (reloc)
> elr: 00000000bff7fde8 lr : 00000000bff7799c
> x0 : 0000000002000f00 x1 : 0000000000000000
> x2 : 00000000fffffffb x3 : 0000000000000020
> x4 : 0000000000000000 x5 : 0000000000000030
> x6 : 0000000000000002 x7 : 00000000bdf4ecf0
> x8 : 0000000000000000 x9 : 0000000000000008
> x10: 00000000ffffffe0 x11: 0000000000000006
> x12: 000000000001869f x13: 0000000000005339
> x14: 0000000000006694 x15: 00000000ffffffff
> x16: 0000000000002080 x17: 0000000000004180
> x18: 00000000bdf58d70 x19: 01e0000002000f00
> x20: 00000000bffec258 x21: 00000000bffcca10
> x22: 01e0000002000f00 x23: 0000000000000006
> x24: 0000000000000000 x25: 0000000000000000
> x26: 0000000000000000 x27: 0000000000000000
> x28: 0000000000000000 x29: 00000000bdf4edc0
>
> Code: b9400aa0 51000400 b9000aa0 aa1603f3 (f94002d6)
> Resetting CPU ...
>
> resetting ...
>
> Reverting just this patch makes it boot fine again.
>
> Cheers
>
> Marcel



-- 
Angelo Dureghello
Timesys
email: angelo.dureghello at timesys.com
cell.:  +39 388 8550663