Fix OEMCrypto documentation for (0,0) patterns

The docs on OEMCrypto_DecryptCENC() weren't updated correctly when we
allowed the (0,0) pattern in v17. This patch brings the header docs
in-line with the handwritten part of the devsite documentation.

Merged from http://go/wvgerrit/204630

Bug: 336330529
Change-Id: Ic6c81d8f04904b83c34fbc0235ebbae8705a4182

libwvdrmengine/oemcrypto/include/OEMCryptoCENC.h

@@ -2237,10 +2237,20 @@ OEMCryptoResult OEMCrypto_GetKeyHandle(OEMCrypto_SESSION session,
  * usually be non-zero. This mode allows devices to decrypt FMP4 HLS content,
  * SAMPLE-AES HLS content, as well as content using the DASH 'cbcs' scheme.
  *
- * The skip field of OEMCrypto_CENCEncryptPatternDesc may also be zero. If
- * the skip field is zero, then patterns are not in use and all crypto blocks
- * in the protected part of the subsample are encrypted. It is not valid for
- * the encrypt field to be zero.
+ * The skip field of OEMCrypto_CENCEncryptPatternDesc may be zero. If the skip
+ * field is zero, then patterns are not in use and all crypto blocks in the
+ * protected part of the subsample are encrypted, except for any partial crypto
+ * blocks at the end. The most common pattern with a skip field of zero is
+ * (10,0), but all patterns with a skip field of zero are functionally the same.
+ *
+ * If the skip field of OEMCrypto_CENCEncryptPatternDesc is zero, the encrypt
+ * field may also be zero. This pattern sometimes appears in content,
+ * particularly in audio tracks. This (0,0) pattern should be treated as
+ * equivalent to the pattern (10,0). e.g. All complete crypto blocks should be
+ * decrypted.
+ *
+ * It is not valid for the encrypt field of OEMCrypto_CENCEncryptPatternDesc to
+ * be zero if the skip field is non-zero.
  *
  * The length of a crypto block in AES-128 is 16 bytes. In the 'cbcs' scheme,
  * if the protected part of a subsample has a length that is not a multiple