| Message ID | 20150310033028.9542.7205.stgit@perseus.themaw.net |
|---|---|
| State | Superseded |
| Delegated to: | Rafał Miłecki |
| Headers | show |
On 10 March 2015 at 04:30, Ian Kent <raven@themaw.net> wrote: > The R8000 nvram is larger than 32k, increase nvram allocation to > 64k to accomadate. I handled this by sending nvram.c patch for mainline: http://patchwork.linux-mips.org/patch/9651/ And backporting all changes to bcm53xx: https://dev.openwrt.org/changeset/45204/
On Thu, 2015-04-02 at 09:45 +0200, Rafał Miłecki wrote: > On 10 March 2015 at 04:30, Ian Kent <raven@themaw.net> wrote: > > The R8000 nvram is larger than 32k, increase nvram allocation to > > 64k to accomadate. > > I handled this by sending nvram.c patch for mainline: > http://patchwork.linux-mips.org/patch/9651/ > And backporting all changes to bcm53xx: > https://dev.openwrt.org/changeset/45204/ OK, help me out with this patch. I thought it was necessary to calculate a from starting offset based on mtd size and nvram size since the code assumes the nvram partition might not start at 0 whereas in the patch it looks like a 0 starting offset is assumed? When I was working on this area of code, IIRC, the from offset was actually not 0. Ian
On Thu, 2015-04-02 at 17:47 +0800, Ian Kent wrote: > On Thu, 2015-04-02 at 09:45 +0200, Rafał Miłecki wrote: > > On 10 March 2015 at 04:30, Ian Kent <raven@themaw.net> wrote: > > > The R8000 nvram is larger than 32k, increase nvram allocation to > > > 64k to accomadate. > > > > I handled this by sending nvram.c patch for mainline: > > http://patchwork.linux-mips.org/patch/9651/ > > And backporting all changes to bcm53xx: > > https://dev.openwrt.org/changeset/45204/ > > OK, help me out with this patch. > > I thought it was necessary to calculate a from starting offset based on > mtd size and nvram size since the code assumes the nvram partition might > not start at 0 whereas in the patch it looks like a 0 starting offset is > assumed? > > When I was working on this area of code, IIRC, the from offset was > actually not 0. Mmm .. so that's probably wrong since ... mtd = get_mtd_device_nm("nvram"); Should get us the start of the nvram partition at offset 0. > > Ian
On 2 April 2015 at 11:57, Ian Kent <raven@themaw.net> wrote: > On Thu, 2015-04-02 at 17:47 +0800, Ian Kent wrote: >> On Thu, 2015-04-02 at 09:45 +0200, Rafał Miłecki wrote: >> > On 10 March 2015 at 04:30, Ian Kent <raven@themaw.net> wrote: >> > > The R8000 nvram is larger than 32k, increase nvram allocation to >> > > 64k to accomadate. >> > >> > I handled this by sending nvram.c patch for mainline: >> > http://patchwork.linux-mips.org/patch/9651/ >> > And backporting all changes to bcm53xx: >> > https://dev.openwrt.org/changeset/45204/ >> >> OK, help me out with this patch. >> >> I thought it was necessary to calculate a from starting offset based on >> mtd size and nvram size since the code assumes the nvram partition might >> not start at 0 whereas in the patch it looks like a 0 starting offset is >> assumed? >> >> When I was working on this area of code, IIRC, the from offset was >> actually not 0. > > Mmm .. so that's probably wrong since ... > > mtd = get_mtd_device_nm("nvram"); > > Should get us the start of the nvram partition at offset 0. That's right, if something doesn't work for you, let me know.
diff --git a/target/linux/bcm53xx/patches-3.14/113-bcm53xx-increase-nvram-allocation-size-to-64k.patch b/target/linux/bcm53xx/patches-3.14/113-bcm53xx-increase-nvram-allocation-size-to-64k.patch new file mode 100644 index 0000000..89763eb --- /dev/null +++ b/target/linux/bcm53xx/patches-3.14/113-bcm53xx-increase-nvram-allocation-size-to-64k.patch @@ -0,0 +1,19 @@ +bcm53xx: increase nvram allocation size to 64k + +The R8000 nvram is too large for the 32k allocation size in +bcm47xx_nvram.c. + +Increase the allocation size to 64k. + +Signed-off-by: Ian Kent <raven@themaw.net> +--- a/drivers/firmware/broadcom/bcm47xx_nvram.c ++++ b/drivers/firmware/broadcom/bcm47xx_nvram.c +@@ -20,7 +20,7 @@ + #include <linux/bcm47xx_nvram.h> + + #define NVRAM_MAGIC 0x48534C46 /* 'FLSH' */ +-#define NVRAM_SPACE 0x8000 ++#define NVRAM_SPACE 0x10000 + #define NVRAM_MAX_GPIO_ENTRIES 32 + #define NVRAM_MAX_GPIO_VALUE_LEN 30 + diff --git a/target/linux/bcm53xx/patches-3.18/113-bcm53xx-increase-nvram-allocation-size-to-64k.patch b/target/linux/bcm53xx/patches-3.18/113-bcm53xx-increase-nvram-allocation-size-to-64k.patch new file mode 100644 index 0000000..3382549 --- /dev/null +++ b/target/linux/bcm53xx/patches-3.18/113-bcm53xx-increase-nvram-allocation-size-to-64k.patch @@ -0,0 +1,20 @@ +bcm53xx: increase nvram allocation size to 64k + +The R8000 nvram is too large for the 32k allocation size in +bcm47xx_nvram.c. + +Increase the allocation size to 64k. + +Signed-off-by: Ian Kent <raven@themaw.net> + +--- a/drivers/firmware/broadcom/bcm47xx_nvram.c ++++ b/drivers/firmware/broadcom/bcm47xx_nvram.c +@@ -20,7 +20,7 @@ + #include <linux/bcm47xx_nvram.h> + + #define NVRAM_MAGIC 0x48534C46 /* 'FLSH' */ +-#define NVRAM_SPACE 0x8000 ++#define NVRAM_SPACE 0x10000 + #define NVRAM_MAX_GPIO_ENTRIES 32 + #define NVRAM_MAX_GPIO_VALUE_LEN 30 +
The R8000 nvram is larger than 32k, increase nvram allocation to 64k to accomadate. Signed-off-by: Ian Kent <raven@themaw.net> --- ...3xx-increase-nvram-allocation-size-to-64k.patch | 19 +++++++++++++++++++ ...3xx-increase-nvram-allocation-size-to-64k.patch | 20 ++++++++++++++++++++ 2 files changed, 39 insertions(+) create mode 100644 target/linux/bcm53xx/patches-3.14/113-bcm53xx-increase-nvram-allocation-size-to-64k.patch create mode 100644 target/linux/bcm53xx/patches-3.18/113-bcm53xx-increase-nvram-allocation-size-to-64k.patch