| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Fri, 1 Oct 2010 13:35:16 -0300 From: Rajiv Andrade <srajiv@...ux.vnet.ibm.com> To: Christophe Henri RICARD <christophe-h.ricard@...com> Cc: James Morris <jmorris@...ei.org>, "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>, "joe@...ches.com" <joe@...ches.com> Subject: Re: [PATCH 1/3] TPM: new stm i2c device driver Hi Christophe, On 01/10/2010, at 12:44, Christophe Henri RICARD wrote: > Hi Rajiv, > > I had implemented the ioctl function with TPMIOC_CANCEL/TPMIOC_TRANSMIT because I have seen that when I was trying to run a TSS tcsd(trousers) process some errors or warning messages were displayed. > I have just checked the trousers source code I have and it use this way in src/tddl/tddl.c . > Trousers supports both methods, device read/write or through ioctl, the warning just notifies that the second isn't going to be used due the device driver's lack of support. > For your information also, the 2 buffers tpm_i2c_buffer instead of the data buffer in my driver was because I found what I suppose to be an error with the allocation of chip->data_buffer in the tpm_open function. > When I have done the following test with the tpm_tis driver for example on a pc platform (x86) with a TPM: > - No TPM application launch such as TSS (trousers). > - Put the platform in standby > - Wake up the platform > dmesg will shows an error during suspend/resume functions execution. > My understanding is that in that specific case, no buffer is allocated and the driver couldn't save the TPM context (TPM_SaveState). Sounds like an issue, thanks for reporting, I'll double check/debug this. > In the case the platform generate a RESET on the LPC bus or on the TPM during standby all the TPM volatile context will be cleared > (all the PCR's for example). I think this could be a problem. > Only for my information, I would like to know if this scenario (TPM RESET during standby) could happen and if it could be a TPM/security issue. As long as it implies in occurring a full power cycle, it isn't a TPM problem. > > I suppose what you are calling security-next tree is linux-next tree on www.kernel.org or are you referring to another tree ? > I will try to provide you a patch for the ioctl function asap. > I'm using this one as security-next: git://git.kernel.org/pub/scm/linux/kernel/git/dhowells/jamesm-security-next.git Thanks, Rajiv > > > -----Original Message----- > From: Rajiv Andrade [mailto:srajiv@...ux.vnet.ibm.com] > Sent: Friday, October 01, 2010 10:26 AM > To: Christophe Henri RICARD > Cc: James Morris; linux-kernel@...r.kernel.org; joe@...ches.com > Subject: Re: [PATCH 1/3] TPM: new stm i2c device driver > > Hi Christophe, > > The code must be applied and compiled against the security-next tree, that currently doesn't provide the locked ioctl anymore. Can you implement the locking inside tpm_st19_i2c_ioctl() and set it at the unlocked_ioctl field? Shouldn't be too hard, you can get tpm_write() as an example. > > By the way, I may be missing something, but why the tpm command transmission (TPMIOC_TRANSMIT) should be handled through ioctl instead of a modified tpm_write() that can send the commands through i2c? > > Thanks, > Rajiv > > On 01/10/2010, at 11:27, Christophe Henri RICARD wrote: > >> Hi James, >> >> Which kernel are you using ? >> I have made my testing with a 2.6.35.4 kernel. >> As far as I know kernel 2.6.36 will not provide any ioctl field any more. >> May be this is the reason ? >> >> Please let me know. >> Thanks >> Christophe >> >> -----Original Message----- >> From: James Morris [mailto:jmorris@...ei.org] >> Sent: Thursday, September 30, 2010 7:51 PM >> To: Rajiv Andrade >> Cc: linux-kernel@...r.kernel.org; joe@...ches.com; Christophe Henri RICARD >> Subject: Re: [PATCH 1/3] TPM: new stm i2c device driver >> >> On Thu, 30 Sep 2010, Rajiv Andrade wrote: >> >>> From: Christophe Henri RICARD <christophe-h.ricard@...com> >>> >>> This driver uses the Linux I2C, TPM and generic gpio interfaces. >>> >>> Signed-off-by: Christophe Henri RICARD <christophe-h.ricard@...com> >>> Signed-off-by: Rajiv Andrade <srajiv@...ux.vnet.ibm.com> >> >> This does not look to have been compile tested against my tree: >> >> CC [M] drivers/char/tpm/tpm_stm_st19_i2c.o >> drivers/char/tpm/tpm_stm_st19_i2c.c:574: error: unknown field ioctl specified in initializer >> drivers/char/tpm/tpm_stm_st19_i2c.c:574: warning: initialization from incompatible pointer type >> make[1]: *** [drivers/char/tpm/tpm_stm_st19_i2c.o] Error 1 >> >> The locked ioctl was removed in commit >> b19dd42faf413b4705d4adb38521e82d73fa4249 >> >> >> >> - James >> -- >> James Morris >> <jmorris@...ei.org> >> -- >> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in >> the body of a message to majordomo@...r.kernel.org >> More majordomo info at http://vger.kernel.org/majordomo-info.html >> Please read the FAQ at http://www.tux.org/lkml/ > -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@...r.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists