On Mon, Dec 19, 2011 at 7:38 PM, Wolfgang Denk <w...@denx.de> wrote: > Dear Che-liang Chiou, > > In message > <canjuy2jq3nf2mjpxrs-8jgksiv1zg2sfcnqdmrksaqpifup...@mail.gmail.com> you > wrote: >> >> > Should we not rather remove all this dead code again? >> > >> > Until today there are no users for this code in mainline, and no >> > patches have been submitted that intend to use it. >> >> Chrome/Chromium OS uses TPM for its secure boot. So I would say it is >> quite a lot of usage on the critical path of booting. > > I do not see any such code in mainline. So for mainline, this is just > dead code, i. e. it adds maintenance efforts without benefit for any > mainline users. > >> The code that uses TPM did not send to the mainline because the >> mainline did not have a TPM driver until very recently. > > Well, the patche shave been submitted long time ago, so I would have > expected to see such code soon after. But so far, nothign happened. > > Now you are trying to add even more code, with still no users in > sight.
The board that I test the TPM driver is Seaboard (a tegra2-based board). The config settings enabling TPM are literally pasted as follows. However, I can't submit the confis settings (so that Seaboard would be the first mainline user of the TPM driver) as part of this patchset because the tegra i2c driver is not yet in the mainline and so enabling TPM will break mainline's Seaboard build. I am 100% certain that tegra i2c driver will be submitted. It just take a few time to polish before sending out for review. And after that I will submit a patch enabling TPM in Seaboard. Would this be sufficient for you? ==== diff --git a/include/configs/seaboard.h b/include/configs/seaboard.h index 7d29144..7c96826 100644 --- a/include/configs/seaboard.h +++ b/include/configs/seaboard.h @@ -42,6 +42,13 @@ #define CONFIG_BOARD_EARLY_INIT_F +/* TPM */ +#define CONFIG_TPM +#define CONFIG_TPM_TIS_I2C +#define CONFIG_TPM_TIS_I2C_BUS_NUMBER 2 +#define CONFIG_TPM_TIS_I2C_SLAVE_ADDRESS 0x20 +#define CONFIG_TPM_TIS_I2C_BURST_LIMITATION 3 + /* SD/MMC */ #define CONFIG_MMC #define CONFIG_GENERIC_MMC ==== >> And, I am still figuring out how to submit the TPM user code. I guess >> it would be better to organize it in a command-line toolkit so that it >> can be interleaved in between other commands. What do you think? > > I don't understand what you mean. Didn't you just write there was > code ready to be submitted? So that code is not ready and not > intended for mainline? I thought you might be interested in Chrome/Chromium OS secure boot, which is implemented as a single monolithic command, and I guess it would be more interested to you if I break it into smaller sub-commands. Anyway, never mind. > Best regards, > > Wolfgang Denk > > -- > DENX Software Engineering GmbH, MD: Wolfgang Denk & Detlev Zundel > HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany > Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: w...@denx.de > Never ascribe to malice that which can adequately be explained by > stupidity. _______________________________________________ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot
