[U-Boot-Users] [PATCH] USB: shutdown USB before booting
This patch fixes a potentially serious issue related to USB which was discouvered by Martin Krause martin.krause@tqs.de and fixed for ARM920T. Martin wrote:
Turn off USB to prevent the host controller from writing to the SDRAM while Linux is booting. This could happen, because the HCCA (Host Controller Communication Area) lies within the SDRAM and the host controller writes continously to this area (as busmaster!), for example to increase the HccaFrameNumber variable, which happens every 1 ms.
This is a slightly modified version of the patch in order to shutdown USB when booting on all architectures.
Signed-off-by: Markus Klotzbuecher mk@denx.de --- common/cmd_bootm.c | 18 ++++++++++++++++++ 1 files changed, 18 insertions(+), 0 deletions(-)
diff --git a/common/cmd_bootm.c b/common/cmd_bootm.c index 959689e..f914815 100644 --- a/common/cmd_bootm.c +++ b/common/cmd_bootm.c @@ -36,6 +36,10 @@ #include <lmb.h> #include <asm/byteorder.h>
+#if (CONFIG_COMMANDS & CFG_CMD_USB) +#include <usb.h> +#endif + #ifdef CFG_HUSH_PARSER #include <hush.h> #endif @@ -213,6 +217,20 @@ int do_bootm (cmd_tbl_t *cmdtp, int flag, int argc, char *argv[]) */ iflag = disable_interrupts();
+#if (CONFIG_COMMANDS & CFG_CMD_USB) + /* + * turn off USB to prevent the host controller from writing to the + * SDRAM while Linux is booting. This could happen (at least for OHCI + * controller), because the HCCA (Host Controller Communication Area) + * lies within the SDRAM and the host controller writes continously to + * this area (as busmaster!). The HccaFrameNumber is for example + * updated every 1 ms within the HCCA structure in SDRAM! For more + * details see the OpenHCI specification. + */ + usb_stop(); +#endif + + #ifdef CONFIG_AMIGAONEG3SE /* * We've possible left the caches enabled during
Hi Markus,
On Thursday 10 July 2008, Markus Klotzbücher wrote:
This patch fixes a potentially serious issue related to USB which was discouvered by Martin Krause martin.krause@tqs.de and fixed for ARM920T. Martin wrote:
Turn off USB to prevent the host controller from writing to the SDRAM while Linux is booting. This could happen, because the HCCA (Host Controller Communication Area) lies within the SDRAM and the host controller writes continously to this area (as busmaster!), for example to increase the HccaFrameNumber variable, which happens every 1 ms.
This is a slightly modified version of the patch in order to shutdown USB when booting on all architectures.
Yes, I remember hearing of this problem. I thought it was already fixed. Seems not to be the case. Please find a comment below.
Signed-off-by: Markus Klotzbuecher mk@denx.de
common/cmd_bootm.c | 18 ++++++++++++++++++ 1 files changed, 18 insertions(+), 0 deletions(-)
diff --git a/common/cmd_bootm.c b/common/cmd_bootm.c
By implementing this usb_stop() in do_bootm() you prevent this problem in many cases (e.g. Linux booting). But unfortunately not in all cases. For example "bootelf" or "bootvx" are not fixed.
Can't this be solved in a different way. By calling usb_stop each time after an USB command completes. IIRC, this is how it is done in the U-Boot network implementation too.
What do you think?
Best regards, Stefan
===================================================================== DENX Software Engineering GmbH, MD: Wolfgang Denk & Detlev Zundel HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany Phone: +49-8142-66989-0 Fax: +49-8142-66989-80 Email: office@denx.de =====================================================================
In message 200807101502.14543.sr@denx.de you wrote:
Yes, I remember hearing of this problem. I thought it was already fixed. Seems not to be the case. Please find a comment below.
I thought so, too. But Martin Krause's patch somehow got stuck in the gearbox.
By implementing this usb_stop() in do_bootm() you prevent this problem in many cases (e.g. Linux booting). But unfortunately not in all cases. For example "bootelf" or "bootvx" are not fixed.
You are right.
Can't this be solved in a different way. By calling usb_stop each time after an USB command completes. IIRC, this is how it is done in the U-Boot network implementation too.
What do you think?
I pulled in the current patch so we reduce the problem as far as possible; any better implementation is welcome.
Best regards,
Wolfgang Denk
In message 87iqvdzz1u.fsf@denx.de you wrote:
This patch fixes a potentially serious issue related to USB which was discouvered by Martin Krause martin.krause@tqs.de and fixed for ARM920T. Martin wrote:
Turn off USB to prevent the host controller from writing to the SDRAM while Linux is booting. This could happen, because the HCCA (Host Controller Communication Area) lies within the SDRAM and the host controller writes continously to this area (as busmaster!), for example to increase the HccaFrameNumber variable, which happens every 1 ms.
This is a slightly modified version of the patch in order to shutdown USB when booting on all architectures.
Signed-off-by: Markus Klotzbuecher mk@denx.de
common/cmd_bootm.c | 18 ++++++++++++++++++ 1 files changed, 18 insertions(+), 0 deletions(-)
Thanks, Markus. This is so urgent that I pull this directly.
Best regards,
Wolfgang Denk
participants (3)
-
Markus Klotzbücher -
Stefan Roese -
Wolfgang Denk