From: Claudio Imbrenda <imbrenda@linux.ibm.com>
To: Janosch Frank <frankja@linux.ibm.com>
Cc: Steffen Eiden <seiden@linux.ibm.com>,
Thomas Huth <thuth@redhat.com>,
David Hildenbrand <david@redhat.com>,
kvm@vger.kernel.org, linux-s390@vger.kernel.org
Subject: Re: [kvm-unit-tests PATCH v3 5/5] s390x: uv-guest: Add attestation tests
Date: Wed, 23 Mar 2022 11:07:28 +0100 [thread overview]
Message-ID: <20220323110728.355735ab@p-imbrenda> (raw)
In-Reply-To: <4396b134-969e-a2f2-2347-cfe5d33925d5@linux.ibm.com>
On Wed, 23 Mar 2022 09:39:27 +0100
Janosch Frank <frankja@linux.ibm.com> wrote:
> On 2/22/22 15:54, Steffen Eiden wrote:
> > Adds several tests to verify correct error paths of attestation.
> >
> > Signed-off-by: Steffen Eiden <seiden@linux.ibm.com>
>
> I think this test deserves its own file: pv-attest.c
sounds like a good idea
>
> But I'd leave the priv check in uv-guest.c.
> @Claudio: Any opinion about having all priv checks here and doing the
> actual execution tests in pv-*.c files?
fine for me
maybe put a comment in pv-attest.c to explain that the priv check is
covered somewhere else already
>
> > ---
> > lib/s390x/asm/uv.h | 5 +-
> > s390x/uv-guest.c | 193 ++++++++++++++++++++++++++++++++++++++++++++-
> > 2 files changed, 196 insertions(+), 2 deletions(-)
> >
> > diff --git a/lib/s390x/asm/uv.h b/lib/s390x/asm/uv.h
> > index c330c0f8..e5f7aa72 100644
> > --- a/lib/s390x/asm/uv.h
> > +++ b/lib/s390x/asm/uv.h
> > @@ -108,7 +108,10 @@ struct uv_cb_qui {
> > u8 reserved88[158 - 136]; /* 0x0088 */
> > uint16_t max_guest_cpus; /* 0x009e */
> > u64 uv_feature_indications; /* 0x00a0 */
> > - u8 reserveda8[200 - 168]; /* 0x00a8 */
> > + uint8_t reserveda8[224 - 168]; /* 0x00a8 */
> > + uint64_t supp_att_hdr_ver; /* 0x00e0 */
> > + uint64_t supp_paf; /* 0x00e8 */
> > + uint8_t reservedf0[256 - 240]; /* 0x00f0 */
> > } __attribute__((packed)) __attribute__((aligned(8)));
> >
> > struct uv_cb_cgc {
> > diff --git a/s390x/uv-guest.c b/s390x/uv-guest.c
> > index 77057bd2..77edbba2 100644
> > --- a/s390x/uv-guest.c
> > +++ b/s390x/uv-guest.c
> > @@ -2,10 +2,11 @@
> > /*
> > * Guest Ultravisor Call tests
> > *
> > - * Copyright (c) 2020 IBM Corp
> > + * Copyright IBM Corp. 2020, 2022
> > *
> > * Authors:
> > * Janosch Frank <frankja@linux.ibm.com>
> > + * Steffen Eiden <seiden@linux.ibm.com>
> > */
> >
> > #include <libcflat.h>
> > @@ -53,6 +54,15 @@ static void test_priv(void)
> > check_pgm_int_code(PGM_INT_CODE_PRIVILEGED_OPERATION);
> > report_prefix_pop();
> >
> > + report_prefix_push("attest");
> > + uvcb.cmd = UVC_CMD_ATTESTATION;
> > + uvcb.len = sizeof(struct uv_cb_attest);
> > + expect_pgm_int();
> > + enter_pstate();
> > + uv_call_once(0, (uint64_t)&uvcb);
> > + check_pgm_int_code(PGM_INT_CODE_PRIVILEGED_OPERATION);
> > + report_prefix_pop();
> > +
> > report_prefix_pop();
> > }
> >
> > @@ -111,7 +121,187 @@ static void test_sharing(void)
> > cc = uv_call(0, (u64)&uvcb);
> > report(cc == 0 && uvcb.header.rc == UVC_RC_EXECUTED, "unshare");
> > report_prefix_pop();
> > +}
> > +
> > +#define ARCB_VERSION_NONE 0
> > +#define ARCB_VERSION_1 0x0100
> > +#define ARCB_MEAS_NONE 0
> > +#define ARCB_MEAS_HMAC_SHA512 1
> > +#define MEASUREMENT_SIZE_HMAC_SHA512 64
> > +#define PAF_PHKH_ATT (1ULL << 61)
> > +#define ADDITIONAL_SIZE_PAF_PHKH_ATT 32
> > +/* arcb with one key slot and no nonce */
> > +struct uv_arcb_v1 {
> > + uint64_t reserved0; /* 0x0000 */
> > + uint32_t req_ver; /* 0x0008 */
> > + uint32_t req_len; /* 0x000c */
> > + uint8_t iv[12]; /* 0x0010 */
> > + uint32_t reserved1c; /* 0x001c */
> > + uint8_t reserved20[7]; /* 0x0020 */
> > + uint8_t nks; /* 0x0027 */
> > + uint32_t reserved28; /* 0x0028 */
> > + uint32_t sea; /* 0x002c */
> > + uint64_t plaint_att_flags; /* 0x0030 */
> > + uint32_t meas_alg_id; /* 0x0038 */
> > + uint32_t reserved3c; /* 0x003c */
> > + uint8_t cpk[160]; /* 0x0040 */
> > + uint8_t key_slot[80]; /* 0x00e0 */
> > + uint8_t meas_key[64]; /* 0x0130 */
> > + uint8_t tag[16]; /* 0x0170 */
> > +} __attribute__((packed));
> > +
> > +struct attest_request_v1 {
> > + struct uv_arcb_v1 arcb;
> > + uint8_t measurement[MEASUREMENT_SIZE_HMAC_SHA512];
> > + uint8_t additional[ADDITIONAL_SIZE_PAF_PHKH_ATT];
> > +};
> > +
> > +static void test_attest_v1(u64 supported_paf)
> > +{
> > + struct uv_cb_attest uvcb = {
> > + .header.cmd = UVC_CMD_ATTESTATION,
> > + .header.len = sizeof(uvcb),
> > + };
> > + struct attest_request_v1 *attest_req = (void *)page;
> > + struct uv_arcb_v1 *arcb = &attest_req->arcb;
> > + int cc;
> > +
> > + memset((void *)page, 0, PAGE_SIZE);
> > +
> > + /*
> > + * Create a minimal arcb/uvcb such that FW has everything to start
> > + * unsealing the request. However, this unsealing will fail as the
> > + * kvm-unit-test framework provides no cryptography functions that
> > + * would be needed to seal such requests.
> > + */
> > + arcb->req_ver = ARCB_VERSION_1;
> > + arcb->req_len = sizeof(*arcb);
> > + arcb->nks = 1;
> > + arcb->sea = sizeof(arcb->meas_key);
> > + arcb->plaint_att_flags = PAF_PHKH_ATT;
> > + arcb->meas_alg_id = ARCB_MEAS_HMAC_SHA512;
> > + uvcb.arcb_addr = (uint64_t)&attest_req->arcb;
> > + uvcb.measurement_address = (uint64_t)attest_req->measurement;
> > + uvcb.measurement_length = sizeof(attest_req->measurement);
> > + uvcb.add_data_address = (uint64_t)attest_req->additional;
> > + uvcb.add_data_length = sizeof(attest_req->additional);
> > +
> > + uvcb.continuation_token = 0xff;
> > + cc = uv_call(0, (uint64_t)&uvcb);
> > + report(cc == 1 && uvcb.header.rc == 0x0101, "invalid continuation token");
> > + uvcb.continuation_token = 0;
> > +
> > + uvcb.user_data_length = sizeof(uvcb.user_data) + 1;
> > + cc = uv_call(0, (uint64_t)&uvcb);
> > + report(cc == 1 && uvcb.header.rc == 0x0102, "invalid user data size");
> > + uvcb.user_data_length = 0;
> > +
> > + uvcb.arcb_addr = get_ram_size() + PAGE_SIZE;
> > + cc = uv_call(0, (uint64_t)&uvcb);
> > + report(cc == 1 && uvcb.header.rc == 0x0103, "invalid address arcb");
> > + uvcb.arcb_addr = page;
> > +
> > + /* 0104 - 0105 need an unseal-able request */
> > +
> > + arcb->req_ver = ARCB_VERSION_NONE;
> > + cc = uv_call(0, (uint64_t)&uvcb);
> > + report(cc == 1 && uvcb.header.rc == 0x0106, "unsupported version");
> > + arcb->req_ver = ARCB_VERSION_1;
> > +
> > + arcb->req_len += 1;
> > + cc = uv_call(0, (uint64_t)&uvcb);
> > + report(cc == 1 && uvcb.header.rc == 0x0107, "invalid arcb size 1");
> > + arcb->req_len -= 1;
> > +
> > + /*
> > + * The arcb needs to grow as well if number of key slots (nks)
> > + * is increased. However, this is not the case and there is no explicit
> > + * 'too many/less nks for that arcb size' error code -> expect 0x0107
> > + */
> > + arcb->nks = 2;
> > + cc = uv_call(0, (uint64_t)&uvcb);
> > + report(cc == 1 && uvcb.header.rc == 0x0107, "invalid arcb size 2");
> > + arcb->nks = 1;
> > +
> > + arcb->nks = 0;
> > + cc = uv_call(0, (uint64_t)&uvcb);
> > + report(cc == 1 && uvcb.header.rc == 0x0108, "invalid num key slots");
> > + arcb->nks = 1;
> > +
> > + /*
> > + * Possible valid size (when using nonce).
> > + * However, req_len too small to host a nonce
> > + */
> > + arcb->sea = 80;
> > + cc = uv_call(0, (uint64_t)&uvcb);
> > + report(cc == 1 && uvcb.header.rc == 0x0109, "invalid encrypted size 1");
> > + arcb->sea = 17;
> > + cc = uv_call(0, (uint64_t)&uvcb);
> > + report(cc == 1 && uvcb.header.rc == 0x0109, "invalid encrypted size 2");
> > + arcb->sea = 64;
> > +
> > + arcb->plaint_att_flags = supported_paf ^ GENMASK_ULL(63, 0);
> > + cc = uv_call(0, (uint64_t)&uvcb);
> > + report(cc == 1 && uvcb.header.rc == 0x010a, "invalid flag");
> > + arcb->plaint_att_flags = PAF_PHKH_ATT;
> > +
> > + arcb->meas_alg_id = ARCB_MEAS_NONE;
> > + cc = uv_call(0, (uint64_t)&uvcb);
> > + report(cc == 1 && uvcb.header.rc == 0x010b, "invalid measurement algorithm");
> > + arcb->meas_alg_id = ARCB_MEAS_HMAC_SHA512;
> >
> > + cc = uv_call(0, (uint64_t)&uvcb);
> > + report(cc == 1 && uvcb.header.rc == 0x010c, "unable unseal");
> > +
> > + uvcb.measurement_length = 0;
> > + cc = uv_call(0, (uint64_t)&uvcb);
> > + report(cc == 1 && uvcb.header.rc == 0x010d, "invalid measurement size");
> > + uvcb.measurement_length = sizeof(attest_req->measurement);
> > +
> > + uvcb.add_data_length = 0;
> > + cc = uv_call(0, (uint64_t)&uvcb);
> > + report(cc == 1 && uvcb.header.rc == 0x010e, "invalid additional size");
> > + uvcb.add_data_length = sizeof(attest_req->additional);
> > +}
> > +
> > +static void test_attest(void)
> > +{
> > + struct uv_cb_attest uvcb = {
> > + .header.cmd = UVC_CMD_ATTESTATION,
> > + .header.len = sizeof(uvcb),
> > + };
> > + const struct uv_cb_qui *uvcb_qui = uv_get_query_data();
> > + int cc;
> > +
> > + report_prefix_push("attest");
> > +
> > + if (!uv_query_test_call(BIT_UVC_CMD_ATTESTATION)) {
> > + report_skip("Attestation not supported.");
> > + goto done;
> > + }
> > +
> > + /* Verify that the UV supports at least one header version */
> > + report(uvcb_qui->supp_att_hdr_ver, "has hdr support");
> > +
> > + memset((void *)page, 0, PAGE_SIZE);
> > +
> > + uvcb.header.len -= 1;
> > + cc = uv_call(0, (uint64_t)&uvcb);
> > + report(cc && uvcb.header.rc == UVC_RC_INV_LEN, "invalid uvcb size 1");
> > + uvcb.header.len += 1;
> > +
> > + uvcb.header.len += 1;
> > + cc = uv_call(0, (uint64_t)&uvcb);
> > + report(cc && uvcb.header.rc == UVC_RC_INV_LEN, "invalid uvcb size 2");
> > + uvcb.header.len -= 1;
> > +
> > + report_prefix_push("v1");
> > + if (test_bit_inv(0, &uvcb_qui->supp_att_hdr_ver))
> > + test_attest_v1(uvcb_qui->supp_paf);
> > + else
> > + report_skip("Attestation version 1 not supported");
> > + report_prefix_pop();
> > +done:
> > report_prefix_pop();
> > }
> >
> > @@ -193,6 +383,7 @@ int main(void)
> > test_invalid();
> > test_query();
> > test_sharing();
> > + test_attest();
> > free_page((void *)page);
> > done:
> > report_prefix_pop();
>
prev parent reply other threads:[~2022-03-23 10:07 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-02-22 14:54 [kvm-unit-tests PATCH v3 0/5] s390x: Attestation tests Steffen Eiden
2022-02-22 14:54 ` [kvm-unit-tests PATCH v3 1/5] s390x: uv-host: Add attestation test Steffen Eiden
2022-03-23 8:18 ` Janosch Frank
2022-02-22 14:54 ` [kvm-unit-tests PATCH v3 2/5] s390x: lib: Add QUI getter Steffen Eiden
2022-02-23 15:37 ` Claudio Imbrenda
2022-03-23 8:07 ` Janosch Frank
2022-02-22 14:54 ` [kvm-unit-tests PATCH v3 3/5] s390x: uv-guest: remove duplicated checks Steffen Eiden
2022-02-22 14:54 ` [kvm-unit-tests PATCH v3 4/5] s390x: uv-guest: add share bit test Steffen Eiden
2022-02-23 15:36 ` Claudio Imbrenda
2022-03-23 8:16 ` Janosch Frank
2022-02-22 14:54 ` [kvm-unit-tests PATCH v3 5/5] s390x: uv-guest: Add attestation tests Steffen Eiden
2022-02-23 15:48 ` Claudio Imbrenda
2022-03-23 8:39 ` Janosch Frank
2022-03-23 10:07 ` Claudio Imbrenda [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20220323110728.355735ab@p-imbrenda \
--to=imbrenda@linux.ibm.com \
--cc=david@redhat.com \
--cc=frankja@linux.ibm.com \
--cc=kvm@vger.kernel.org \
--cc=linux-s390@vger.kernel.org \
--cc=seiden@linux.ibm.com \
--cc=thuth@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox