Hi Patrick, > Thanks Patrice, > > On 4/20/21 12:24 PM, Patrice CHOTARD wrote: > > Hi Patrick > > > > On 3/4/21 5:47 PM, Patrick Delaunay wrote: > >> For NOR devices the logical DFU buffer size is the sector_size, > >> as it is done in dfu_sf.c or in spi/sf_mtd.c > >> (sf_mtd_info.erasesize = flash->sector_size) > >> > >> For NAND the DFU size was already limited to erasesize as > >> has_pages = true. > >> > >> So the mtd dfu backend can use this erasesize for all the MTD > >> devices, NOR and NAND with dfu->max_buf_size = mtd->erasesize > >> > >> This difference was initially copied from MTD command, where > >> data is fully available in RAM without size limitation. > >> > >> This patch avoids to have many sector write in dfu_mtd.c at the end > >> of the DFU transfer and avoids issues with USB timeout or WATCHDOG. > >> > >> Signed-off-by: Patrick Delaunay <patrick.delau...@foss.st.com> > >> --- > >> > >> drivers/dfu/dfu_mtd.c | 5 +---- > >> 1 file changed, 1 insertion(+), 4 deletions(-) > >> > >> diff --git a/drivers/dfu/dfu_mtd.c b/drivers/dfu/dfu_mtd.c > >> index ca67585a7e..7efb3cbd79 100644 > >> --- a/drivers/dfu/dfu_mtd.c > >> +++ b/drivers/dfu/dfu_mtd.c > >> @@ -252,7 +252,6 @@ int dfu_fill_entity_mtd(struct dfu_entity > >> *dfu, char *devstr, char *s) { > >> char *st; > >> struct mtd_info *mtd; > >> - bool has_pages; > >> int ret, part; > >> > >> mtd = get_mtd_device_nm(devstr); > >> @@ -262,9 +261,7 @@ int dfu_fill_entity_mtd(struct dfu_entity > >> *dfu, char *devstr, char *s) > >> dfu->dev_type = DFU_DEV_MTD; > >> dfu->data.mtd.info = mtd; > >> - > >> - has_pages = mtd->type == MTD_NANDFLASH || mtd->type == > >> MTD_MLCNANDFLASH; > >> - dfu->max_buf_size = has_pages ? mtd->erasesize : 0; > >> + dfu->max_buf_size = mtd->erasesize; > >> > >> st = strsep(&s, " "); > >> if (!strcmp(st, "raw")) { > >> > > Reviewed-by: Patrice Chotard <patrice.chot...@foss.st.com> > > > > Thanks > > > > Patrice > > > Lukasz, any remarks on your side? > > > Can I take this patch in my first STM32 pull request for v2020.10 ?
Please pull it to your tree. > > or a PR is already planed on your side ? > > http://patchwork.ozlabs.org/project/uboot/patch/20210304174748.1.I1d824180669d63e22275f082e7ac0fe50c2b8646@changeid/ > > > Regards > > Patrick > Best regards, Lukasz Majewski -- DENX Software Engineering GmbH, Managing Director: Wolfgang Denk HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany Phone: (+49)-8142-66989-59 Fax: (+49)-8142-66989-80 Email: lu...@denx.de
pgpqO11KnJEbw.pgp
Description: OpenPGP digital signature