[U-Boot] [PATCH] cmd_fdt: save fdtaddr in hex format
Commit 90fbee3e4051 ("cmd_fdt: Actually fix fdt command in sandbox") changed the format(from hex address to unsigned long) in which "fdtaddr" is saved . However do_fdt continues reads the "fdtaddr" assuming it to be in hex format. This may lead to fdt being either loaded or attempted to load at erroneous address generating fault if the address is out of memory.
This patch changes back the format to hex while saving the "fdtaddr" as it was done before.
Cc: Linus Walleij linus.walleij@linaro.org Cc: Tom Rini trini@konsulko.com Cc: Simon Glass sjg@chromium.org Cc: Hua Yanghao huayanghao@gmail.com Cc: Heiko Schocher hs@denx.de --- common/cmd_fdt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
Hi,
You will see my signoff missing and it's intentional as I officially can't contribute to Uboot. I found this issue and it's been there for a while now, so thought of atleast reporting it(with a patch). So if anyone can review and submit this or any appropriate fix that would be great. I see this issue on vexpress TC2.
Regards, Sudeep
diff --git a/common/cmd_fdt.c b/common/cmd_fdt.c index 682b6553958f..4c18962d8532 100644 --- a/common/cmd_fdt.c +++ b/common/cmd_fdt.c @@ -45,7 +45,7 @@ void set_working_fdt_addr(ulong addr)
buf = map_sysmem(addr, 0); working_fdt = buf; - setenv_ulong("fdtaddr", addr); + setenv_hex("fdtaddr", addr); }
/* -- 1.9.1
Hi Sudeep,
On 3 July 2015 at 11:29, Sudeep Holla sudeep.holla@arm.com wrote:
Commit 90fbee3e4051 ("cmd_fdt: Actually fix fdt command in sandbox") changed the format(from hex address to unsigned long) in which "fdtaddr" is saved . However do_fdt continues reads the "fdtaddr" assuming it to be in hex format. This may lead to fdt being either loaded or attempted to load at erroneous address generating fault if the address is out of memory.
This patch changes back the format to hex while saving the "fdtaddr" as it was done before.
Cc: Linus Walleij linus.walleij@linaro.org Cc: Tom Rini trini@konsulko.com Cc: Simon Glass sjg@chromium.org Cc: Hua Yanghao huayanghao@gmail.com Cc: Heiko Schocher hs@denx.de
common/cmd_fdt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
Hi,
You will see my signoff missing and it's intentional as I officially can't contribute to Uboot. I found this issue and it's been there for a while now, so thought of atleast reporting it(with a patch). So if anyone can review and submit this or any appropriate fix that would be great. I see this issue on vexpress TC2.
We can't take this without a sign-off - what stops you from adding it exactly?
The patch looks correct to me.
Regards, Simon
Regards, Sudeep
diff --git a/common/cmd_fdt.c b/common/cmd_fdt.c index 682b6553958f..4c18962d8532 100644 --- a/common/cmd_fdt.c +++ b/common/cmd_fdt.c @@ -45,7 +45,7 @@ void set_working_fdt_addr(ulong addr)
buf = map_sysmem(addr, 0); working_fdt = buf;
setenv_ulong("fdtaddr", addr);
setenv_hex("fdtaddr", addr);}
/*
1.9.1
Hi Simon,
On 04/07/15 00:06, Simon Glass wrote:
Hi Sudeep,
On 3 July 2015 at 11:29, Sudeep Holla sudeep.holla@arm.com wrote:
Commit 90fbee3e4051 ("cmd_fdt: Actually fix fdt command in sandbox") changed the format(from hex address to unsigned long) in which "fdtaddr" is saved . However do_fdt continues reads the "fdtaddr" assuming it to be in hex format. This may lead to fdt being either loaded or attempted to load at erroneous address generating fault if the address is out of memory.
This patch changes back the format to hex while saving the "fdtaddr" as it was done before.
Cc: Linus Walleij linus.walleij@linaro.org Cc: Tom Rini trini@konsulko.com Cc: Simon Glass sjg@chromium.org Cc: Hua Yanghao huayanghao@gmail.com Cc: Heiko Schocher hs@denx.de
common/cmd_fdt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
Hi,
You will see my signoff missing and it's intentional as I officially can't contribute to Uboot. I found this issue and it's been there for a while now, so thought of atleast reporting it(with a patch). So if anyone can review and submit this or any appropriate fix that would be great. I see this issue on vexpress TC2.
We can't take this without a sign-off - what stops you from adding it exactly?
Unfortunately it's not in the list of legally approved projects I can contribute. So I am bound not to do so legally.
Hi Linus,
Since you did a lot work on vexpress/Juno quite recently on ARM's request, can you cook up a similar patch to fix the above mentioned issue if that's OK for you ?
Regards, Sudeep
Commit 90fbee3e4051 ("cmd_fdt: Actually fix fdt command in sandbox") changed the format(from hex address to unsigned long) in which "fdtaddr" is saved . However do_fdt continues reads the "fdtaddr" assuming it to be in hex format. This may lead to fdt being either loaded or attempted to load at erroneous address generating fault if the address is out of memory.
This patch changes back the format to hex while saving the "fdtaddr" as it was done before.
Signed-off-by: Sudeep Holla sudeep.holla@arm.com Cc: Linus Walleij linus.walleij@linaro.org Cc: Tom Rini trini@konsulko.com Cc: Simon Glass sjg@chromium.org Cc: Hua Yanghao huayanghao@gmail.com Cc: Heiko Schocher hs@denx.de --- common/cmd_fdt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
Reposting this time with my signoff as it turns out I can legally do so.
diff --git a/common/cmd_fdt.c b/common/cmd_fdt.c index 682b6553958f..4c18962d8532 100644 --- a/common/cmd_fdt.c +++ b/common/cmd_fdt.c @@ -45,7 +45,7 @@ void set_working_fdt_addr(ulong addr)
buf = map_sysmem(addr, 0); working_fdt = buf; - setenv_ulong("fdtaddr", addr); + setenv_hex("fdtaddr", addr); }
/*
On 10 July 2015 at 10:18, Sudeep Holla sudeep.holla@arm.com wrote:
Commit 90fbee3e4051 ("cmd_fdt: Actually fix fdt command in sandbox") changed the format(from hex address to unsigned long) in which "fdtaddr" is saved . However do_fdt continues reads the "fdtaddr" assuming it to be in hex format. This may lead to fdt being either loaded or attempted to load at erroneous address generating fault if the address is out of memory.
This patch changes back the format to hex while saving the "fdtaddr" as it was done before.
Signed-off-by: Sudeep Holla sudeep.holla@arm.com Cc: Linus Walleij linus.walleij@linaro.org Cc: Tom Rini trini@konsulko.com Cc: Simon Glass sjg@chromium.org Cc: Hua Yanghao huayanghao@gmail.com Cc: Heiko Schocher hs@denx.de
common/cmd_fdt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
Reposting this time with my signoff as it turns out I can legally do so.
Great!
- Simon
On Fri, Jul 10, 2015 at 6:18 PM, Sudeep Holla sudeep.holla@arm.com wrote:
Commit 90fbee3e4051 ("cmd_fdt: Actually fix fdt command in sandbox") changed the format(from hex address to unsigned long) in which "fdtaddr" is saved . However do_fdt continues reads the "fdtaddr" assuming it to be in hex format. This may lead to fdt being either loaded or attempted to load at erroneous address generating fault if the address is out of memory.
This patch changes back the format to hex while saving the "fdtaddr" as it was done before.
Signed-off-by: Sudeep Holla sudeep.holla@arm.com Cc: Linus Walleij linus.walleij@linaro.org Cc: Tom Rini trini@konsulko.com Cc: Simon Glass sjg@chromium.org Cc: Hua Yanghao huayanghao@gmail.com Cc: Heiko Schocher hs@denx.de
Looks like the right fix: Acked-by: Linus Walleij linus.walleij@linaro.org
Yours, Linus Walleij
On 16 July 2015 at 07:26, Linus Walleij linus.walleij@linaro.org wrote:
On Fri, Jul 10, 2015 at 6:18 PM, Sudeep Holla sudeep.holla@arm.com wrote:
Commit 90fbee3e4051 ("cmd_fdt: Actually fix fdt command in sandbox") changed the format(from hex address to unsigned long) in which "fdtaddr" is saved . However do_fdt continues reads the "fdtaddr" assuming it to be in hex format. This may lead to fdt being either loaded or attempted to load at erroneous address generating fault if the address is out of memory.
This patch changes back the format to hex while saving the "fdtaddr" as it was done before.
Signed-off-by: Sudeep Holla sudeep.holla@arm.com Cc: Linus Walleij linus.walleij@linaro.org Cc: Tom Rini trini@konsulko.com Cc: Simon Glass sjg@chromium.org Cc: Hua Yanghao huayanghao@gmail.com Cc: Heiko Schocher hs@denx.de
Looks like the right fix: Acked-by: Linus Walleij linus.walleij@linaro.org
Yours, Linus Walleij
Acked-by: Simon Glass sjg@chromium.org
On 16 July 2015 at 10:34, Simon Glass sjg@chromium.org wrote:
On 16 July 2015 at 07:26, Linus Walleij linus.walleij@linaro.org wrote:
On Fri, Jul 10, 2015 at 6:18 PM, Sudeep Holla sudeep.holla@arm.com wrote:
Commit 90fbee3e4051 ("cmd_fdt: Actually fix fdt command in sandbox") changed the format(from hex address to unsigned long) in which "fdtaddr" is saved . However do_fdt continues reads the "fdtaddr" assuming it to be in hex format. This may lead to fdt being either loaded or attempted to load at erroneous address generating fault if the address is out of memory.
This patch changes back the format to hex while saving the "fdtaddr" as it was done before.
Signed-off-by: Sudeep Holla sudeep.holla@arm.com Cc: Linus Walleij linus.walleij@linaro.org Cc: Tom Rini trini@konsulko.com Cc: Simon Glass sjg@chromium.org Cc: Hua Yanghao huayanghao@gmail.com Cc: Heiko Schocher hs@denx.de
Looks like the right fix: Acked-by: Linus Walleij linus.walleij@linaro.org
Yours, Linus Walleij
Acked-by: Simon Glass sjg@chromium.org
Applied to u-boot-fdt, thanks!
participants (3)
-
Linus Walleij -
Simon Glass -
Sudeep Holla