minio/internal/crypto/sse-kms.go

238 lines
8.4 KiB
Go
Raw Normal View History

// Copyright (c) 2015-2021 MinIO, Inc.
//
// This file is part of MinIO Object Storage stack
//
// This program is free software: you can redistribute it and/or modify
// it under the terms of the GNU Affero General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
//
// This program is distributed in the hope that it will be useful
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU Affero General Public License for more details.
//
// You should have received a copy of the GNU Affero General Public License
// along with this program. If not, see <http://www.gnu.org/licenses/>.
package crypto
import (
"context"
"encoding/base64"
"errors"
"net/http"
"path"
"strings"
jsoniter "github.com/json-iterator/go"
xhttp "github.com/minio/minio/internal/http"
"github.com/minio/minio/internal/kms"
"github.com/minio/minio/internal/logger"
)
type ssekms struct{}
var (
// S3KMS represents AWS SSE-KMS. It provides functionality to
// handle SSE-KMS requests.
S3KMS = ssekms{}
_ Type = S3KMS
)
// String returns the SSE domain as string. For SSE-KMS the
// domain is "SSE-KMS".
func (ssekms) String() string { return "SSE-KMS" }
// IsRequested returns true if the HTTP headers contains
// at least one SSE-KMS header.
func (ssekms) IsRequested(h http.Header) bool {
if _, ok := h[xhttp.AmzServerSideEncryptionKmsID]; ok {
return true
}
if _, ok := h[xhttp.AmzServerSideEncryptionKmsContext]; ok {
return true
}
if _, ok := h[xhttp.AmzServerSideEncryption]; ok {
// Return only true if the SSE header is specified and does not contain the SSE-S3 value
return strings.ToUpper(h.Get(xhttp.AmzServerSideEncryption)) != xhttp.AmzEncryptionAES
}
return false
}
// ParseHTTP parses the SSE-KMS headers and returns the SSE-KMS key ID
// and the KMS context on success.
func (ssekms) ParseHTTP(h http.Header) (string, kms.Context, error) {
if h == nil {
return "", nil, ErrInvalidEncryptionMethod
}
algorithm := h.Get(xhttp.AmzServerSideEncryption)
if algorithm != xhttp.AmzEncryptionKMS {
return "", nil, ErrInvalidEncryptionMethod
}
var ctx kms.Context
if context, ok := h[xhttp.AmzServerSideEncryptionKmsContext]; ok {
b, err := base64.StdEncoding.DecodeString(context[0])
if err != nil {
return "", nil, err
}
json := jsoniter.ConfigCompatibleWithStandardLibrary
if err := json.Unmarshal(b, &ctx); err != nil {
return "", nil, err
}
}
keyID := h.Get(xhttp.AmzServerSideEncryptionKmsID)
spaces := strings.HasPrefix(keyID, " ") || strings.HasSuffix(keyID, " ")
if spaces {
return "", nil, ErrInvalidEncryptionKeyID
}
return strings.TrimPrefix(keyID, ARNPrefix), ctx, nil
}
// IsEncrypted returns true if the object metadata indicates
// that the object was uploaded using SSE-KMS.
func (ssekms) IsEncrypted(metadata map[string]string) bool {
if _, ok := metadata[MetaSealedKeyKMS]; ok {
return true
}
return false
}
// UnsealObjectKey extracts and decrypts the sealed object key
// from the metadata using KMS and returns the decrypted object
// key.
func (s3 ssekms) UnsealObjectKey(k kms.KMS, metadata map[string]string, bucket, object string) (key ObjectKey, err error) {
if k == nil {
return key, Errorf("KMS not configured")
}
crypto: add support for decrypting SSE-KMS metadata (#11415) This commit refactors the SSE implementation and add S3-compatible SSE-KMS context handling. SSE-KMS differs from SSE-S3 in two main aspects: 1. The client can request a particular key and specify a KMS context as part of the request. 2. The ETag of an SSE-KMS encrypted object is not the MD5 sum of the object content. This commit only focuses on the 1st aspect. A client can send an optional SSE context when using SSE-KMS. This context is remembered by the S3 server such that the client does not have to specify the context again (during multipart PUT / GET / HEAD ...). The crypto. context also includes the bucket/object name to prevent renaming objects at the backend. Now, AWS S3 behaves as following: - If the user does not provide a SSE-KMS context it does not store one - resp. does not include the SSE-KMS context header in the response (e.g. HEAD). - If the user specifies a SSE-KMS context without the bucket/object name then AWS stores the exact context the client provided but adds the bucket/object name internally. The response contains the KMS context without the bucket/object name. - If the user specifies a SSE-KMS context with the bucket/object name then AWS again stores the exact context provided by the client. The response contains the KMS context with the bucket/object name. This commit implements this behavior w.r.t. SSE-KMS. However, as of now, no such object can be created since the server rejects SSE-KMS encryption requests. This commit is one stepping stone for SSE-KMS support. Co-authored-by: Harshavardhana <harsha@minio.io>
2021-02-03 18:19:08 -05:00
keyID, kmsKey, sealedKey, ctx, err := s3.ParseMetadata(metadata)
if err != nil {
return key, err
}
if ctx == nil {
ctx = kms.Context{bucket: path.Join(bucket, object)}
} else if _, ok := ctx[bucket]; !ok {
crypto: add support for decrypting SSE-KMS metadata (#11415) This commit refactors the SSE implementation and add S3-compatible SSE-KMS context handling. SSE-KMS differs from SSE-S3 in two main aspects: 1. The client can request a particular key and specify a KMS context as part of the request. 2. The ETag of an SSE-KMS encrypted object is not the MD5 sum of the object content. This commit only focuses on the 1st aspect. A client can send an optional SSE context when using SSE-KMS. This context is remembered by the S3 server such that the client does not have to specify the context again (during multipart PUT / GET / HEAD ...). The crypto. context also includes the bucket/object name to prevent renaming objects at the backend. Now, AWS S3 behaves as following: - If the user does not provide a SSE-KMS context it does not store one - resp. does not include the SSE-KMS context header in the response (e.g. HEAD). - If the user specifies a SSE-KMS context without the bucket/object name then AWS stores the exact context the client provided but adds the bucket/object name internally. The response contains the KMS context without the bucket/object name. - If the user specifies a SSE-KMS context with the bucket/object name then AWS again stores the exact context provided by the client. The response contains the KMS context with the bucket/object name. This commit implements this behavior w.r.t. SSE-KMS. However, as of now, no such object can be created since the server rejects SSE-KMS encryption requests. This commit is one stepping stone for SSE-KMS support. Co-authored-by: Harshavardhana <harsha@minio.io>
2021-02-03 18:19:08 -05:00
ctx[bucket] = path.Join(bucket, object)
}
unsealKey, err := k.DecryptKey(keyID, kmsKey, ctx)
if err != nil {
return key, err
}
err = key.Unseal(unsealKey, sealedKey, s3.String(), bucket, object)
return key, err
}
// CreateMetadata encodes the sealed object key into the metadata and returns
// the modified metadata. If the keyID and the kmsKey is not empty it encodes
// both into the metadata as well. It allocates a new metadata map if metadata
// is nil.
func (ssekms) CreateMetadata(metadata map[string]string, keyID string, kmsKey []byte, sealedKey SealedKey, ctx kms.Context) map[string]string {
if sealedKey.Algorithm != SealAlgorithm {
logger.CriticalIf(context.Background(), Errorf("The seal algorithm '%s' is invalid for SSE-S3", sealedKey.Algorithm))
}
// There are two possibilites:
// - We use a KMS -> There must be non-empty key ID and a KMS data key.
// - We use a K/V -> There must be no key ID and no KMS data key.
// Otherwise, the caller has passed an invalid argument combination.
if keyID == "" && len(kmsKey) != 0 {
logger.CriticalIf(context.Background(), errors.New("The key ID must not be empty if a KMS data key is present"))
}
if keyID != "" && len(kmsKey) == 0 {
logger.CriticalIf(context.Background(), errors.New("The KMS data key must not be empty if a key ID is present"))
}
if metadata == nil {
metadata = make(map[string]string, 5)
}
metadata[MetaAlgorithm] = sealedKey.Algorithm
metadata[MetaIV] = base64.StdEncoding.EncodeToString(sealedKey.IV[:])
metadata[MetaSealedKeyKMS] = base64.StdEncoding.EncodeToString(sealedKey.Key[:])
if len(ctx) > 0 {
b, _ := ctx.MarshalText()
metadata[MetaContext] = base64.StdEncoding.EncodeToString(b)
}
if len(kmsKey) > 0 && keyID != "" { // We use a KMS -> Store key ID and sealed KMS data key.
metadata[MetaKeyID] = keyID
metadata[MetaDataEncryptionKey] = base64.StdEncoding.EncodeToString(kmsKey)
}
return metadata
}
// ParseMetadata extracts all SSE-KMS related values from the object metadata
// and checks whether they are well-formed. It returns the sealed object key
// on success. If the metadata contains both, a KMS master key ID and a sealed
// KMS data key it returns both. If the metadata does not contain neither a
// KMS master key ID nor a sealed KMS data key it returns an empty keyID and
// KMS data key. Otherwise, it returns an error.
func (ssekms) ParseMetadata(metadata map[string]string) (keyID string, kmsKey []byte, sealedKey SealedKey, ctx kms.Context, err error) {
// Extract all required values from object metadata
b64IV, ok := metadata[MetaIV]
if !ok {
crypto: add support for decrypting SSE-KMS metadata (#11415) This commit refactors the SSE implementation and add S3-compatible SSE-KMS context handling. SSE-KMS differs from SSE-S3 in two main aspects: 1. The client can request a particular key and specify a KMS context as part of the request. 2. The ETag of an SSE-KMS encrypted object is not the MD5 sum of the object content. This commit only focuses on the 1st aspect. A client can send an optional SSE context when using SSE-KMS. This context is remembered by the S3 server such that the client does not have to specify the context again (during multipart PUT / GET / HEAD ...). The crypto. context also includes the bucket/object name to prevent renaming objects at the backend. Now, AWS S3 behaves as following: - If the user does not provide a SSE-KMS context it does not store one - resp. does not include the SSE-KMS context header in the response (e.g. HEAD). - If the user specifies a SSE-KMS context without the bucket/object name then AWS stores the exact context the client provided but adds the bucket/object name internally. The response contains the KMS context without the bucket/object name. - If the user specifies a SSE-KMS context with the bucket/object name then AWS again stores the exact context provided by the client. The response contains the KMS context with the bucket/object name. This commit implements this behavior w.r.t. SSE-KMS. However, as of now, no such object can be created since the server rejects SSE-KMS encryption requests. This commit is one stepping stone for SSE-KMS support. Co-authored-by: Harshavardhana <harsha@minio.io>
2021-02-03 18:19:08 -05:00
return keyID, kmsKey, sealedKey, ctx, errMissingInternalIV
}
algorithm, ok := metadata[MetaAlgorithm]
if !ok {
crypto: add support for decrypting SSE-KMS metadata (#11415) This commit refactors the SSE implementation and add S3-compatible SSE-KMS context handling. SSE-KMS differs from SSE-S3 in two main aspects: 1. The client can request a particular key and specify a KMS context as part of the request. 2. The ETag of an SSE-KMS encrypted object is not the MD5 sum of the object content. This commit only focuses on the 1st aspect. A client can send an optional SSE context when using SSE-KMS. This context is remembered by the S3 server such that the client does not have to specify the context again (during multipart PUT / GET / HEAD ...). The crypto. context also includes the bucket/object name to prevent renaming objects at the backend. Now, AWS S3 behaves as following: - If the user does not provide a SSE-KMS context it does not store one - resp. does not include the SSE-KMS context header in the response (e.g. HEAD). - If the user specifies a SSE-KMS context without the bucket/object name then AWS stores the exact context the client provided but adds the bucket/object name internally. The response contains the KMS context without the bucket/object name. - If the user specifies a SSE-KMS context with the bucket/object name then AWS again stores the exact context provided by the client. The response contains the KMS context with the bucket/object name. This commit implements this behavior w.r.t. SSE-KMS. However, as of now, no such object can be created since the server rejects SSE-KMS encryption requests. This commit is one stepping stone for SSE-KMS support. Co-authored-by: Harshavardhana <harsha@minio.io>
2021-02-03 18:19:08 -05:00
return keyID, kmsKey, sealedKey, ctx, errMissingInternalSealAlgorithm
}
b64SealedKey, ok := metadata[MetaSealedKeyKMS]
if !ok {
crypto: add support for decrypting SSE-KMS metadata (#11415) This commit refactors the SSE implementation and add S3-compatible SSE-KMS context handling. SSE-KMS differs from SSE-S3 in two main aspects: 1. The client can request a particular key and specify a KMS context as part of the request. 2. The ETag of an SSE-KMS encrypted object is not the MD5 sum of the object content. This commit only focuses on the 1st aspect. A client can send an optional SSE context when using SSE-KMS. This context is remembered by the S3 server such that the client does not have to specify the context again (during multipart PUT / GET / HEAD ...). The crypto. context also includes the bucket/object name to prevent renaming objects at the backend. Now, AWS S3 behaves as following: - If the user does not provide a SSE-KMS context it does not store one - resp. does not include the SSE-KMS context header in the response (e.g. HEAD). - If the user specifies a SSE-KMS context without the bucket/object name then AWS stores the exact context the client provided but adds the bucket/object name internally. The response contains the KMS context without the bucket/object name. - If the user specifies a SSE-KMS context with the bucket/object name then AWS again stores the exact context provided by the client. The response contains the KMS context with the bucket/object name. This commit implements this behavior w.r.t. SSE-KMS. However, as of now, no such object can be created since the server rejects SSE-KMS encryption requests. This commit is one stepping stone for SSE-KMS support. Co-authored-by: Harshavardhana <harsha@minio.io>
2021-02-03 18:19:08 -05:00
return keyID, kmsKey, sealedKey, ctx, Errorf("The object metadata is missing the internal sealed key for SSE-S3")
}
// There are two possibilites:
// - We use a KMS -> There must be a key ID and a KMS data key.
// - We use a K/V -> There must be no key ID and no KMS data key.
// Otherwise, the metadata is corrupted.
keyID, idPresent := metadata[MetaKeyID]
b64KMSSealedKey, kmsKeyPresent := metadata[MetaDataEncryptionKey]
if !idPresent && kmsKeyPresent {
crypto: add support for decrypting SSE-KMS metadata (#11415) This commit refactors the SSE implementation and add S3-compatible SSE-KMS context handling. SSE-KMS differs from SSE-S3 in two main aspects: 1. The client can request a particular key and specify a KMS context as part of the request. 2. The ETag of an SSE-KMS encrypted object is not the MD5 sum of the object content. This commit only focuses on the 1st aspect. A client can send an optional SSE context when using SSE-KMS. This context is remembered by the S3 server such that the client does not have to specify the context again (during multipart PUT / GET / HEAD ...). The crypto. context also includes the bucket/object name to prevent renaming objects at the backend. Now, AWS S3 behaves as following: - If the user does not provide a SSE-KMS context it does not store one - resp. does not include the SSE-KMS context header in the response (e.g. HEAD). - If the user specifies a SSE-KMS context without the bucket/object name then AWS stores the exact context the client provided but adds the bucket/object name internally. The response contains the KMS context without the bucket/object name. - If the user specifies a SSE-KMS context with the bucket/object name then AWS again stores the exact context provided by the client. The response contains the KMS context with the bucket/object name. This commit implements this behavior w.r.t. SSE-KMS. However, as of now, no such object can be created since the server rejects SSE-KMS encryption requests. This commit is one stepping stone for SSE-KMS support. Co-authored-by: Harshavardhana <harsha@minio.io>
2021-02-03 18:19:08 -05:00
return keyID, kmsKey, sealedKey, ctx, Errorf("The object metadata is missing the internal KMS key-ID for SSE-S3")
}
if idPresent && !kmsKeyPresent {
crypto: add support for decrypting SSE-KMS metadata (#11415) This commit refactors the SSE implementation and add S3-compatible SSE-KMS context handling. SSE-KMS differs from SSE-S3 in two main aspects: 1. The client can request a particular key and specify a KMS context as part of the request. 2. The ETag of an SSE-KMS encrypted object is not the MD5 sum of the object content. This commit only focuses on the 1st aspect. A client can send an optional SSE context when using SSE-KMS. This context is remembered by the S3 server such that the client does not have to specify the context again (during multipart PUT / GET / HEAD ...). The crypto. context also includes the bucket/object name to prevent renaming objects at the backend. Now, AWS S3 behaves as following: - If the user does not provide a SSE-KMS context it does not store one - resp. does not include the SSE-KMS context header in the response (e.g. HEAD). - If the user specifies a SSE-KMS context without the bucket/object name then AWS stores the exact context the client provided but adds the bucket/object name internally. The response contains the KMS context without the bucket/object name. - If the user specifies a SSE-KMS context with the bucket/object name then AWS again stores the exact context provided by the client. The response contains the KMS context with the bucket/object name. This commit implements this behavior w.r.t. SSE-KMS. However, as of now, no such object can be created since the server rejects SSE-KMS encryption requests. This commit is one stepping stone for SSE-KMS support. Co-authored-by: Harshavardhana <harsha@minio.io>
2021-02-03 18:19:08 -05:00
return keyID, kmsKey, sealedKey, ctx, Errorf("The object metadata is missing the internal sealed KMS data key for SSE-S3")
}
// Check whether all extracted values are well-formed
iv, err := base64.StdEncoding.DecodeString(b64IV)
if err != nil || len(iv) != 32 {
crypto: add support for decrypting SSE-KMS metadata (#11415) This commit refactors the SSE implementation and add S3-compatible SSE-KMS context handling. SSE-KMS differs from SSE-S3 in two main aspects: 1. The client can request a particular key and specify a KMS context as part of the request. 2. The ETag of an SSE-KMS encrypted object is not the MD5 sum of the object content. This commit only focuses on the 1st aspect. A client can send an optional SSE context when using SSE-KMS. This context is remembered by the S3 server such that the client does not have to specify the context again (during multipart PUT / GET / HEAD ...). The crypto. context also includes the bucket/object name to prevent renaming objects at the backend. Now, AWS S3 behaves as following: - If the user does not provide a SSE-KMS context it does not store one - resp. does not include the SSE-KMS context header in the response (e.g. HEAD). - If the user specifies a SSE-KMS context without the bucket/object name then AWS stores the exact context the client provided but adds the bucket/object name internally. The response contains the KMS context without the bucket/object name. - If the user specifies a SSE-KMS context with the bucket/object name then AWS again stores the exact context provided by the client. The response contains the KMS context with the bucket/object name. This commit implements this behavior w.r.t. SSE-KMS. However, as of now, no such object can be created since the server rejects SSE-KMS encryption requests. This commit is one stepping stone for SSE-KMS support. Co-authored-by: Harshavardhana <harsha@minio.io>
2021-02-03 18:19:08 -05:00
return keyID, kmsKey, sealedKey, ctx, errInvalidInternalIV
}
if algorithm != SealAlgorithm {
crypto: add support for decrypting SSE-KMS metadata (#11415) This commit refactors the SSE implementation and add S3-compatible SSE-KMS context handling. SSE-KMS differs from SSE-S3 in two main aspects: 1. The client can request a particular key and specify a KMS context as part of the request. 2. The ETag of an SSE-KMS encrypted object is not the MD5 sum of the object content. This commit only focuses on the 1st aspect. A client can send an optional SSE context when using SSE-KMS. This context is remembered by the S3 server such that the client does not have to specify the context again (during multipart PUT / GET / HEAD ...). The crypto. context also includes the bucket/object name to prevent renaming objects at the backend. Now, AWS S3 behaves as following: - If the user does not provide a SSE-KMS context it does not store one - resp. does not include the SSE-KMS context header in the response (e.g. HEAD). - If the user specifies a SSE-KMS context without the bucket/object name then AWS stores the exact context the client provided but adds the bucket/object name internally. The response contains the KMS context without the bucket/object name. - If the user specifies a SSE-KMS context with the bucket/object name then AWS again stores the exact context provided by the client. The response contains the KMS context with the bucket/object name. This commit implements this behavior w.r.t. SSE-KMS. However, as of now, no such object can be created since the server rejects SSE-KMS encryption requests. This commit is one stepping stone for SSE-KMS support. Co-authored-by: Harshavardhana <harsha@minio.io>
2021-02-03 18:19:08 -05:00
return keyID, kmsKey, sealedKey, ctx, errInvalidInternalSealAlgorithm
}
encryptedKey, err := base64.StdEncoding.DecodeString(b64SealedKey)
if err != nil || len(encryptedKey) != 64 {
crypto: add support for decrypting SSE-KMS metadata (#11415) This commit refactors the SSE implementation and add S3-compatible SSE-KMS context handling. SSE-KMS differs from SSE-S3 in two main aspects: 1. The client can request a particular key and specify a KMS context as part of the request. 2. The ETag of an SSE-KMS encrypted object is not the MD5 sum of the object content. This commit only focuses on the 1st aspect. A client can send an optional SSE context when using SSE-KMS. This context is remembered by the S3 server such that the client does not have to specify the context again (during multipart PUT / GET / HEAD ...). The crypto. context also includes the bucket/object name to prevent renaming objects at the backend. Now, AWS S3 behaves as following: - If the user does not provide a SSE-KMS context it does not store one - resp. does not include the SSE-KMS context header in the response (e.g. HEAD). - If the user specifies a SSE-KMS context without the bucket/object name then AWS stores the exact context the client provided but adds the bucket/object name internally. The response contains the KMS context without the bucket/object name. - If the user specifies a SSE-KMS context with the bucket/object name then AWS again stores the exact context provided by the client. The response contains the KMS context with the bucket/object name. This commit implements this behavior w.r.t. SSE-KMS. However, as of now, no such object can be created since the server rejects SSE-KMS encryption requests. This commit is one stepping stone for SSE-KMS support. Co-authored-by: Harshavardhana <harsha@minio.io>
2021-02-03 18:19:08 -05:00
return keyID, kmsKey, sealedKey, ctx, Errorf("The internal sealed key for SSE-KMS is invalid")
}
if idPresent && kmsKeyPresent { // We are using a KMS -> parse the sealed KMS data key.
kmsKey, err = base64.StdEncoding.DecodeString(b64KMSSealedKey)
if err != nil {
crypto: add support for decrypting SSE-KMS metadata (#11415) This commit refactors the SSE implementation and add S3-compatible SSE-KMS context handling. SSE-KMS differs from SSE-S3 in two main aspects: 1. The client can request a particular key and specify a KMS context as part of the request. 2. The ETag of an SSE-KMS encrypted object is not the MD5 sum of the object content. This commit only focuses on the 1st aspect. A client can send an optional SSE context when using SSE-KMS. This context is remembered by the S3 server such that the client does not have to specify the context again (during multipart PUT / GET / HEAD ...). The crypto. context also includes the bucket/object name to prevent renaming objects at the backend. Now, AWS S3 behaves as following: - If the user does not provide a SSE-KMS context it does not store one - resp. does not include the SSE-KMS context header in the response (e.g. HEAD). - If the user specifies a SSE-KMS context without the bucket/object name then AWS stores the exact context the client provided but adds the bucket/object name internally. The response contains the KMS context without the bucket/object name. - If the user specifies a SSE-KMS context with the bucket/object name then AWS again stores the exact context provided by the client. The response contains the KMS context with the bucket/object name. This commit implements this behavior w.r.t. SSE-KMS. However, as of now, no such object can be created since the server rejects SSE-KMS encryption requests. This commit is one stepping stone for SSE-KMS support. Co-authored-by: Harshavardhana <harsha@minio.io>
2021-02-03 18:19:08 -05:00
return keyID, kmsKey, sealedKey, ctx, Errorf("The internal sealed KMS data key for SSE-KMS is invalid")
}
}
b64Ctx, ok := metadata[MetaContext]
if ok {
b, err := base64.StdEncoding.DecodeString(b64Ctx)
if err != nil {
return keyID, kmsKey, sealedKey, ctx, Errorf("The internal KMS context is not base64-encoded")
}
json := jsoniter.ConfigCompatibleWithStandardLibrary
if err = json.Unmarshal(b, &ctx); err != nil {
return keyID, kmsKey, sealedKey, ctx, Errorf("The internal sealed KMS context is invalid %w", err)
}
}
sealedKey.Algorithm = algorithm
copy(sealedKey.IV[:], iv)
copy(sealedKey.Key[:], encryptedKey)
crypto: add support for decrypting SSE-KMS metadata (#11415) This commit refactors the SSE implementation and add S3-compatible SSE-KMS context handling. SSE-KMS differs from SSE-S3 in two main aspects: 1. The client can request a particular key and specify a KMS context as part of the request. 2. The ETag of an SSE-KMS encrypted object is not the MD5 sum of the object content. This commit only focuses on the 1st aspect. A client can send an optional SSE context when using SSE-KMS. This context is remembered by the S3 server such that the client does not have to specify the context again (during multipart PUT / GET / HEAD ...). The crypto. context also includes the bucket/object name to prevent renaming objects at the backend. Now, AWS S3 behaves as following: - If the user does not provide a SSE-KMS context it does not store one - resp. does not include the SSE-KMS context header in the response (e.g. HEAD). - If the user specifies a SSE-KMS context without the bucket/object name then AWS stores the exact context the client provided but adds the bucket/object name internally. The response contains the KMS context without the bucket/object name. - If the user specifies a SSE-KMS context with the bucket/object name then AWS again stores the exact context provided by the client. The response contains the KMS context with the bucket/object name. This commit implements this behavior w.r.t. SSE-KMS. However, as of now, no such object can be created since the server rejects SSE-KMS encryption requests. This commit is one stepping stone for SSE-KMS support. Co-authored-by: Harshavardhana <harsha@minio.io>
2021-02-03 18:19:08 -05:00
return keyID, kmsKey, sealedKey, ctx, nil
}