[PATCH 0/3] mtd: fix AMD/Intel flash bugs

Joakim Tjernlund Joakim.Tjernlund at infinera.com
Mon Mar 12 04:11:04 PDT 2018


On Mon, 2018-03-12 at 10:09 +0100, Andrea Adami wrote:
> CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and know the content is safe.
> 
> 
> On Sun, Mar 11, 2018 at 5:06 PM, Joakim Tjernlund
> <Joakim.Tjernlund at infinera.com> wrote:
> > On Thu, 2018-03-01 at 14:39 +0100, Joakim Tjernlund wrote:
> > > This fixes a bug which allows read/write to a erase block currently
> > > erasing.
> > > Also works around a chip bug in Micron 28F00AP30 chips.
> > > 
> > > Joakim Tjernlund (3):
> > >   cfi_cmdset_0001: Do not allow read/write to suspend erase block.
> > >   cfi_cmdset_0001: Workaround Micron Erase suspend bug.
> > >   cfi_cmdset_0002: Do not allow read/write to suspend erase block.
> > > 
> > >  drivers/mtd/chips/cfi_cmdset_0001.c | 33 ++++++++++++++++++++++++++++-----
> > >  drivers/mtd/chips/cfi_cmdset_0002.c |  9 ++++++---
> > >  include/linux/mtd/flashchip.h       |  1 +
> > >  3 files changed, 35 insertions(+), 8 deletions(-)
> > > 
> > 
> > Ping?
> > This is a fairly serious bug as it confuses the flash state machine,
> > making further flash access fail.
> > 
> 
> Hello,
> 
> I don't undertsand exactly from the titles what is happening.
> Do you want to disable erase and wite suspend like in this other 2016
> patch found around?

Not quite, the first part is about preventing read/write to a block that is
being erased. This can happen if two apps opens the same /dev/mtd device.

Then there is a bug fix for erase suspend not working in the smaller boot block
sections, that part is similar to the patch you reference but much less invasive

> https://patchwork.kernel.org/patch/9380029/
> 
> See, I am user of the Sharp LH28F640BF NOR, very similar to the Intel
> Startaflash, 4-Plane Page Mode Dual Work.
> On these chips erase suspend works just fine.
> What has been disabled is dual work: simultaneous read-while-erase and
> read-while-program because there has never been real support for
> n-planes in kernel.
> 
> What is exactly the issue?
> Regards
> Andrea
> 
> >  Jocke
> > ______________________________________________________
> > Linux MTD discussion mailing list
> > http://lists.infradead.org/mailman/listinfo/linux-mtd/


More information about the linux-mtd mailing list