linux-kernel - Re: [PATCH v38 14/24] x86/sgx: Add SGX_IOC_ENCLAVE

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

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20200923144707.GF5160@linux.intel.com>
Date:   Wed, 23 Sep 2020 17:47:07 +0300
From:   Jarkko Sakkinen <jarkko.sakkinen@...ux.intel.com>
To:     Borislav Petkov <bp@...en8.de>
Cc:     Jethro Beekman <jethro@...tanix.com>,
        Sean Christopherson <sean.j.christopherson@...el.com>,
        x86@...nel.org, linux-sgx@...r.kernel.org,
        linux-kernel@...r.kernel.org,
        Haitao Huang <haitao.huang@...ux.intel.com>,
        Chunyang Hui <sanqian.hcy@...fin.com>,
        Jordan Hand <jorhand@...ux.microsoft.com>,
        Nathaniel McCallum <npmccallum@...hat.com>,
        Seth Moore <sethmo@...gle.com>,
        Darren Kenny <darren.kenny@...cle.com>,
        Suresh Siddha <suresh.b.siddha@...el.com>,
        akpm@...ux-foundation.org, andriy.shevchenko@...ux.intel.com,
        asapek@...gle.com, cedric.xing@...el.com, chenalexchen@...gle.com,
        conradparker@...gle.com, cyhanish@...gle.com,
        dave.hansen@...el.com, haitao.huang@...el.com,
        josh@...htriplett.org, kai.huang@...el.com, kai.svahn@...el.com,
        kmoy@...gle.com, ludloff@...gle.com, luto@...nel.org,
        nhorman@...hat.com, puiterwijk@...hat.com, rientjes@...gle.com,
        tglx@...utronix.de, yaozhangx@...gle.com
Subject: Re: [PATCH v38 14/24] x86/sgx: Add SGX_IOC_ENCLAVE_INIT

On Tue, Sep 22, 2020 at 04:29:09PM +0200, Borislav Petkov wrote:
> On Tue, Sep 22, 2020 at 02:56:19PM +0200, Jethro Beekman wrote:
> > I don't see why you'd need to retry indefinitely. Yes the MSRs may not
> > match the cached value for “reasons”, but if after you've written
> > them once it still doesn't work, clearly either 1) an “unhelpful”
> > VMM is actively messing with the MSRs which I'd say is at best a VMM
> > bug or 2) there was an EPC reset and your enclave is now invalid
> > anyway, so no need to EINIT.
> 
> /me likes that even more.

OK, so I did not follow this particular discussion in high detail, so
as a sanity check I'll preview my changes.

I'd refine sgx_update_lepubkeyhash_msrs() to:

static void sgx_update_lepubkeyhash_msrs(u64 *lepubkeyhash)
{
	u64 *cache;
	int i;

	preempt_disable();

	cache = per_cpu(sgx_lepubkeyhash_cache, smp_processor_id());
	for (i = 0; i < 4; i++) {
		if (lepubkeyhash[i] != cache[i]) {
			wrmsrl(MSR_IA32_SGXLEPUBKEYHASH0 + i, lepubkeyhash[i]);
			cache[i] = lepubkeyhash[i];
		}
	}

	preempt_enable();
}

I'd drop sgx_einit() completely.

Finally, in sgx_encl_init() I would call sgx_update_lepubkeyhash_msrs()
directly:

/* ... */
			sgx_update_lepubkeyhash_msrs(mrsigner);

			ret = __einit(sigstruct, token, sgx_get_epc_addr(secs));
/* ... */

These staments would replace the call to sgx_einit().

Do I have the correct understanding?

/Jarkko