diff --git a/ctl.go b/ctl.go
index 9709e40..3e90951 100644
--- a/ctl.go
+++ b/ctl.go
@@ -800,7 +800,7 @@ func servectlcmd(ctx context.Context, ctl *ctl, shutdown func()) {
 							m.Size = correctSize
 
 							mr := acc.MessageReader(m)
-							part, err := message.EnsurePart(mr, m.Size)
+							part, err := message.EnsurePart(log, false, mr, m.Size)
 							if err != nil {
 								_, werr := fmt.Fprintf(w, "parsing message %d again: %v (continuing)\n", m.ID, err)
 								ctl.xcheck(werr, "write")
@@ -890,7 +890,7 @@ func servectlcmd(ctx context.Context, ctl *ctl, shutdown func()) {
 					return q.ForEach(func(m store.Message) error {
 						lastID = m.ID
 						mr := acc.MessageReader(m)
-						p, err := message.EnsurePart(mr, m.Size)
+						p, err := message.EnsurePart(log, false, mr, m.Size)
 						if err != nil {
 							_, err := fmt.Fprintf(w, "parsing message %d: %v (continuing)\n", m.ID, err)
 							ctl.xcheck(err, "write")
diff --git a/dmarcrpt/parse.go b/dmarcrpt/parse.go
index 5106065..a727bb3 100644
--- a/dmarcrpt/parse.go
+++ b/dmarcrpt/parse.go
@@ -13,6 +13,7 @@ import (
 	"strings"
 
 	"github.com/mjl-/mox/message"
+	"github.com/mjl-/mox/mlog"
 	"github.com/mjl-/mox/moxio"
 )
 
@@ -33,17 +34,17 @@ func ParseReport(r io.Reader) (*Feedback, error) {
 // ParseMessageReport parses an aggregate feedback report from a mail message. The
 // maximum message size is 15MB, the maximum report size after decompression is
 // 20MB.
-func ParseMessageReport(r io.ReaderAt) (*Feedback, error) {
+func ParseMessageReport(log *mlog.Log, r io.ReaderAt) (*Feedback, error) {
 	// ../rfc/7489:1801
-	p, err := message.Parse(&moxio.LimitAtReader{R: r, Limit: 15 * 1024 * 1024})
+	p, err := message.Parse(log, true, &moxio.LimitAtReader{R: r, Limit: 15 * 1024 * 1024})
 	if err != nil {
 		return nil, fmt.Errorf("parsing mail message: %s", err)
 	}
 
-	return parseMessageReport(p)
+	return parseMessageReport(log, p)
 }
 
-func parseMessageReport(p message.Part) (*Feedback, error) {
+func parseMessageReport(log *mlog.Log, p message.Part) (*Feedback, error) {
 	// Pretty much any mime structure is allowed. ../rfc/7489:1861
 	// In practice, some parties will send the report as the only (non-multipart)
 	// content of the message.
@@ -53,14 +54,14 @@ func parseMessageReport(p message.Part) (*Feedback, error) {
 	}
 
 	for {
-		sp, err := p.ParseNextPart()
+		sp, err := p.ParseNextPart(log)
 		if err == io.EOF {
 			return nil, ErrNoReport
 		}
 		if err != nil {
 			return nil, err
 		}
-		report, err := parseMessageReport(*sp)
+		report, err := parseMessageReport(log, *sp)
 		if err == ErrNoReport {
 			continue
 		} else if err != nil || report != nil {
diff --git a/dmarcrpt/parse_test.go b/dmarcrpt/parse_test.go
index 96f1ed1..a93d3c0 100644
--- a/dmarcrpt/parse_test.go
+++ b/dmarcrpt/parse_test.go
@@ -5,8 +5,12 @@ import (
 	"reflect"
 	"strings"
 	"testing"
+
+	"github.com/mjl-/mox/mlog"
 )
 
+var xlog = mlog.New("dmarcrpt")
+
 const reportExample = `<?xml version="1.0" encoding="UTF-8" ?>
 <feedback>
   <report_metadata>
@@ -130,7 +134,7 @@ func TestParseMessageReport(t *testing.T) {
 		if err != nil {
 			t.Fatalf("open %q: %s", p, err)
 		}
-		_, err = ParseMessageReport(f)
+		_, err = ParseMessageReport(xlog, f)
 		if err != nil {
 			t.Fatalf("ParseMessageReport: %q: %s", p, err)
 		}
@@ -138,7 +142,7 @@ func TestParseMessageReport(t *testing.T) {
 	}
 
 	// No report in a non-multipart message.
-	_, err = ParseMessageReport(strings.NewReader("From: <mjl@mox.example>\r\n\r\nNo report.\r\n"))
+	_, err = ParseMessageReport(xlog, strings.NewReader("From: <mjl@mox.example>\r\n\r\nNo report.\r\n"))
 	if err != ErrNoReport {
 		t.Fatalf("message without report, got err %#v, expected ErrNoreport", err)
 	}
@@ -164,7 +168,7 @@ MIME-Version: 1.0
 
 --===============5735553800636657282==--
 `, "\n", "\r\n")
-	_, err = ParseMessageReport(strings.NewReader(multipartNoreport))
+	_, err = ParseMessageReport(xlog, strings.NewReader(multipartNoreport))
 	if err != ErrNoReport {
 		t.Fatalf("message without report, got err %#v, expected ErrNoreport", err)
 	}
diff --git a/dsn/dsn_test.go b/dsn/dsn_test.go
index 9d431eb..01bda47 100644
--- a/dsn/dsn_test.go
+++ b/dsn/dsn_test.go
@@ -19,6 +19,8 @@ import (
 	"github.com/mjl-/mox/smtp"
 )
 
+var xlog = mlog.New("dsn")
+
 func xparseDomain(s string) dns.Domain {
 	d, err := dns.ParseDomain(s)
 	if err != nil {
@@ -33,7 +35,7 @@ func xparseIPDomain(s string) dns.IPDomain {
 
 func tparseMessage(t *testing.T, data []byte, nparts int) (*Message, *message.Part) {
 	t.Helper()
-	m, p, err := Parse(bytes.NewReader(data))
+	m, p, err := Parse(xlog, bytes.NewReader(data))
 	if err != nil {
 		t.Fatalf("parsing dsn: %v", err)
 	}
diff --git a/dsn/parse.go b/dsn/parse.go
index c2ab695..161508b 100644
--- a/dsn/parse.go
+++ b/dsn/parse.go
@@ -11,6 +11,7 @@ import (
 
 	"github.com/mjl-/mox/dns"
 	"github.com/mjl-/mox/message"
+	"github.com/mjl-/mox/mlog"
 	"github.com/mjl-/mox/smtp"
 )
 
@@ -22,17 +23,17 @@ import (
 // The first return value is the machine-parsed DSN message. The second value is
 // the entire MIME multipart message. Use its Parts field to access the
 // human-readable text and optional original message/headers.
-func Parse(r io.ReaderAt) (*Message, *message.Part, error) {
+func Parse(log *mlog.Log, r io.ReaderAt) (*Message, *message.Part, error) {
 	// DSNs can mix and match subtypes with and without utf-8. ../rfc/6533:441
 
-	part, err := message.Parse(r)
+	part, err := message.Parse(log, false, r)
 	if err != nil {
 		return nil, nil, fmt.Errorf("parsing message: %v", err)
 	}
 	if part.MediaType != "MULTIPART" || part.MediaSubType != "REPORT" {
 		return nil, nil, fmt.Errorf(`message has content-type %q, must have "message/report"`, strings.ToLower(part.MediaType+"/"+part.MediaSubType))
 	}
-	err = part.Walk(nil)
+	err = part.Walk(log, nil)
 	if err != nil {
 		return nil, nil, fmt.Errorf("parsing message parts: %v", err)
 	}
diff --git a/imapserver/append_test.go b/imapserver/append_test.go
index d74b408..fbde472 100644
--- a/imapserver/append_test.go
+++ b/imapserver/append_test.go
@@ -55,7 +55,7 @@ func TestAppend(t *testing.T) {
 	tc3.transactf("ok", "noop")
 	tc3.xuntagged() // Inbox is not selected, nothing to report.
 
-	tc2.transactf("ok", "append inbox (\\Seen) \" 1-Jan-2022 10:10:00 +0100\" UTF8 ({34+}\r\ncontent-type: text/plain;;\r\n\r\ntest)")
+	tc2.transactf("ok", "append inbox (\\Seen) \" 1-Jan-2022 10:10:00 +0100\" UTF8 ({47+}\r\ncontent-type: just completely invalid;;\r\n\r\ntest)")
 	tc2.xuntagged(imapclient.UntaggedExists(2))
 	tc2.xcodeArg(imapclient.CodeAppendUID{UIDValidity: 1, UID: 2})
 
diff --git a/import.go b/import.go
index 5e8f878..b3d42fa 100644
--- a/import.go
+++ b/import.go
@@ -325,7 +325,7 @@ func importctl(ctx context.Context, ctl *ctl, mbox bool) {
 			mb.Add(m.MailboxCounts())
 
 			// Parse message and store parsed information for later fast retrieval.
-			p, err := message.EnsurePart(msgf, m.Size)
+			p, err := message.EnsurePart(ctl.log, false, msgf, m.Size)
 			if err != nil {
 				ctl.log.Infox("parsing message, continuing", err, mlog.Field("path", origPath))
 			}
diff --git a/junk.go b/junk.go
index 83ba273..409a9e3 100644
--- a/junk.go
+++ b/junk.go
@@ -309,6 +309,8 @@ func cmdJunkPlay(c *cmd) {
 
 	var nbad, nnodate, nham, nspam, nsent int
 
+	jlog := mlog.New("junkplay")
+
 	scanDir := func(dir string, ham, sent bool) {
 		for _, name := range listDir(dir) {
 			path := dir + "/" + name
@@ -316,7 +318,7 @@ func cmdJunkPlay(c *cmd) {
 			xcheckf(err, "open %q", path)
 			fi, err := mf.Stat()
 			xcheckf(err, "stat %q", path)
-			p, err := message.EnsurePart(mf, fi.Size())
+			p, err := message.EnsurePart(jlog, false, mf, fi.Size())
 			if err != nil {
 				nbad++
 				if err := mf.Close(); err != nil {
@@ -396,7 +398,7 @@ func cmdJunkPlay(c *cmd) {
 			}()
 			fi, err := mf.Stat()
 			xcheckf(err, "stat %q", path)
-			p, err := message.EnsurePart(mf, fi.Size())
+			p, err := message.EnsurePart(jlog, false, mf, fi.Size())
 			if err != nil {
 				log.Printf("bad sent message %q: %s", path, err)
 				return
diff --git a/junk/filter.go b/junk/filter.go
index afceeee..b96a21f 100644
--- a/junk/filter.go
+++ b/junk/filter.go
@@ -501,7 +501,7 @@ func (f *Filter) ClassifyMessagePath(ctx context.Context, path string) (probabil
 }
 
 func (f *Filter) ClassifyMessageReader(ctx context.Context, mf io.ReaderAt, size int64) (probability float64, words map[string]struct{}, nham, nspam int, rerr error) {
-	m, err := message.EnsurePart(mf, size)
+	m, err := message.EnsurePart(f.log, false, mf, size)
 	if err != nil && errors.Is(err, message.ErrBadContentType) {
 		// Invalid content-type header is a sure sign of spam.
 		//f.log.Infox("parsing content", err)
@@ -567,7 +567,7 @@ func (f *Filter) Train(ctx context.Context, ham bool, words map[string]struct{})
 }
 
 func (f *Filter) TrainMessage(ctx context.Context, r io.ReaderAt, size int64, ham bool) error {
-	p, _ := message.EnsurePart(r, size)
+	p, _ := message.EnsurePart(f.log, false, r, size)
 	words, err := f.ParseMessage(p)
 	if err != nil {
 		return fmt.Errorf("parsing mail contents: %v", err)
@@ -576,7 +576,7 @@ func (f *Filter) TrainMessage(ctx context.Context, r io.ReaderAt, size int64, ha
 }
 
 func (f *Filter) UntrainMessage(ctx context.Context, r io.ReaderAt, size int64, ham bool) error {
-	p, _ := message.EnsurePart(r, size)
+	p, _ := message.EnsurePart(f.log, false, r, size)
 	words, err := f.ParseMessage(p)
 	if err != nil {
 		return fmt.Errorf("parsing mail contents: %v", err)
diff --git a/junk/parse.go b/junk/parse.go
index c480444..d5b8859 100644
--- a/junk/parse.go
+++ b/junk/parse.go
@@ -31,7 +31,7 @@ func (f *Filter) tokenizeMail(path string) (bool, map[string]struct{}, error) {
 	if err != nil {
 		return false, nil, err
 	}
-	p, _ := message.EnsurePart(mf, fi.Size())
+	p, _ := message.EnsurePart(f.log, false, mf, fi.Size())
 	words, err := f.ParseMessage(p)
 	return true, words, err
 }
diff --git a/main.go b/main.go
index 9cf7acf..e4903a4 100644
--- a/main.go
+++ b/main.go
@@ -1443,11 +1443,13 @@ headers prepended.
 		c.Usage()
 	}
 
+	clog := mlog.New("dkimsign")
+
 	msgf, err := os.Open(args[0])
 	xcheckf(err, "open message")
 	defer msgf.Close()
 
-	p, err := message.Parse(msgf)
+	p, err := message.Parse(clog, true, msgf)
 	xcheckf(err, "parsing message")
 
 	if len(p.Envelope.From) != 1 {
@@ -1591,7 +1593,7 @@ can be found in message headers.
 
 	data, err := io.ReadAll(os.Stdin)
 	xcheckf(err, "read message")
-	dmarcFrom, _, err := message.From(bytes.NewReader(data))
+	dmarcFrom, _, err := message.From(mlog.New("dmarcverify"), false, bytes.NewReader(data))
 	xcheckf(err, "extract dmarc from message")
 
 	const ignoreTestMode = false
@@ -1621,10 +1623,12 @@ understand email deliverability problems.
 		c.Usage()
 	}
 
+	clog := mlog.New("dmarcparsereportmsg")
+
 	for _, arg := range args {
 		f, err := os.Open(arg)
 		xcheckf(err, "open %q", arg)
-		feedback, err := dmarcrpt.ParseMessageReport(f)
+		feedback, err := dmarcrpt.ParseMessageReport(clog, f)
 		xcheckf(err, "parse report in %q", arg)
 		meta := feedback.ReportMetadata
 		fmt.Printf("Report: period %s-%s, organisation %q, reportID %q, %s\n", time.Unix(meta.DateRange.Begin, 0).UTC().String(), time.Unix(meta.DateRange.End, 0).UTC().String(), meta.OrgName, meta.ReportID, meta.Email)
@@ -1673,9 +1677,11 @@ func cmdDMARCDBAddReport(c *cmd) {
 
 	mustLoadConfig()
 
+	clog := mlog.New("dmarcdbaddreport")
+
 	fromdomain := xparseDomain(args[0], "domain")
 	fmt.Fprintln(os.Stderr, "reading report message from stdin")
-	report, err := dmarcrpt.ParseMessageReport(os.Stdin)
+	report, err := dmarcrpt.ParseMessageReport(clog, os.Stdin)
 	xcheckf(err, "parse message")
 	err = dmarcdb.AddReport(context.Background(), report, fromdomain)
 	xcheckf(err, "add dmarc report")
@@ -1712,10 +1718,12 @@ The report is printed in formatted JSON.
 		c.Usage()
 	}
 
+	clog := mlog.New("tlsrptparsereportmsg")
+
 	for _, arg := range args {
 		f, err := os.Open(arg)
 		xcheckf(err, "open %q", arg)
-		report, err := tlsrpt.ParseMessage(f)
+		report, err := tlsrpt.ParseMessage(clog, f)
 		xcheckf(err, "parse report in %q", arg)
 		// todo future: only print the highlights?
 		enc := json.NewEncoder(os.Stdout)
@@ -1855,11 +1863,13 @@ func cmdTLSRPTDBAddReport(c *cmd) {
 
 	mustLoadConfig()
 
+	clog := mlog.New("tlsrptdbaddreport")
+
 	// First read message, to get the From-header. Then parse it as TLSRPT.
 	fmt.Fprintln(os.Stderr, "reading report message from stdin")
 	buf, err := io.ReadAll(os.Stdin)
 	xcheckf(err, "reading message")
-	part, err := message.Parse(bytes.NewReader(buf))
+	part, err := message.Parse(clog, true, bytes.NewReader(buf))
 	xcheckf(err, "parsing message")
 	if part.Envelope == nil || len(part.Envelope.From) != 1 {
 		log.Fatalf("message must have one From-header")
@@ -1867,7 +1877,7 @@ func cmdTLSRPTDBAddReport(c *cmd) {
 	from := part.Envelope.From[0]
 	domain := xparseDomain(from.Host, "domain")
 
-	report, err := tlsrpt.ParseMessage(bytes.NewReader(buf))
+	report, err := tlsrpt.ParseMessage(clog, bytes.NewReader(buf))
 	xcheckf(err, "parsing tls report in message")
 
 	mailfrom := from.User + "@" + from.Host // todo future: should escape and such
@@ -2276,6 +2286,8 @@ func cmdEnsureParsed(c *cmd) {
 		c.Usage()
 	}
 
+	clog := mlog.New("ensureparsed")
+
 	mustLoadConfig()
 	a, err := store.OpenAccount(args[0])
 	xcheckf(err, "open account")
@@ -2298,7 +2310,7 @@ func cmdEnsureParsed(c *cmd) {
 		}
 		for _, m := range l {
 			mr := a.MessageReader(m)
-			p, err := message.EnsurePart(mr, m.Size)
+			p, err := message.EnsurePart(clog, false, mr, m.Size)
 			if err != nil {
 				log.Printf("parsing message %d: %v (continuing)", m.ID, err)
 			}
@@ -2351,13 +2363,15 @@ func cmdMessageParse(c *cmd) {
 		c.Usage()
 	}
 
+	clog := mlog.New("messageparse")
+
 	f, err := os.Open(args[0])
 	xcheckf(err, "open")
 	defer f.Close()
 
-	part, err := message.Parse(f)
+	part, err := message.Parse(clog, false, f)
 	xcheckf(err, "parsing message")
-	err = part.Walk(nil)
+	err = part.Walk(clog, nil)
 	xcheckf(err, "parsing nested parts")
 	enc := json.NewEncoder(os.Stdout)
 	enc.SetIndent("", "\t")
diff --git a/message/from.go b/message/from.go
index 9550479..66a3c19 100644
--- a/message/from.go
+++ b/message/from.go
@@ -6,6 +6,7 @@ import (
 	"net/textproto"
 
 	"github.com/mjl-/mox/dns"
+	"github.com/mjl-/mox/mlog"
 	"github.com/mjl-/mox/smtp"
 )
 
@@ -16,12 +17,12 @@ import (
 // From headers may be present. From returns an error if there is not exactly
 // one address. This address can be used for evaluating a DMARC policy against
 // SPF and DKIM results.
-func From(r io.ReaderAt) (raddr smtp.Address, header textproto.MIMEHeader, rerr error) {
+func From(log *mlog.Log, strict bool, r io.ReaderAt) (raddr smtp.Address, header textproto.MIMEHeader, rerr error) {
 	// ../rfc/7489:1243
 
 	// todo: only allow utf8 if enabled in session/message?
 
-	p, err := Parse(r)
+	p, err := Parse(log, strict, r)
 	if err != nil {
 		// todo: should we continue with p, perhaps headers can be parsed?
 		return raddr, nil, fmt.Errorf("parsing message: %v", err)
diff --git a/message/part.go b/message/part.go
index f193bc5..b5894a8 100644
--- a/message/part.go
+++ b/message/part.go
@@ -7,7 +7,6 @@ package message
 // todo: should we allow base64 messages where a line starts with a space? and possibly more whitespace. is happening in messages. coreutils base64 accepts it, encoding/base64 does not.
 // todo: handle comments in headers?
 // todo: should we just always store messages with \n instead of \r\n? \r\n seems easier for use with imap.
-// todo: is a header always \r\n\r\n-separated? or is \r\n enough at the beginning of a file? because what would this mean: "\r\ndata"? data isn't a header.
 // todo: can use a cleanup
 
 import (
@@ -32,8 +31,6 @@ import (
 	"github.com/mjl-/mox/smtp"
 )
 
-var xlog = mlog.New("message")
-
 var (
 	ErrBadContentType = errors.New("bad content-type")
 )
@@ -83,6 +80,7 @@ type Part struct {
 	lastBoundOffset int64                // Start of header of last/previous part. Used to skip a part if ParseNextPart is called and nextBoundOffset is -1.
 	parent          *Part                // Parent part, for getting bound from, and setting nextBoundOffset when a part has finished reading. Only for subparts, not top-level parts.
 	bound           []byte               // Only set if valid multipart with boundary, includes leading --, excludes \r\n.
+	strict          bool                 // If set, valid crlf line endings are verified when reading body.
 }
 
 // todo: have all Content* fields in Part?
@@ -112,18 +110,24 @@ type Address struct {
 
 // Parse reads the headers of the mail message and returns a part.
 // A part provides access to decoded and raw contents of a message and its multiple parts.
-func Parse(r io.ReaderAt) (Part, error) {
-	return newPart(r, 0, nil)
+//
+// If strict is set, fewer attempts are made to continue parsing when errors are
+// encountered, such as with invalid content-type headers or bare carriage returns.
+func Parse(log *mlog.Log, strict bool, r io.ReaderAt) (Part, error) {
+	return newPart(log, strict, r, 0, nil)
 }
 
 // EnsurePart parses a part as with Parse, but ensures a usable part is always
 // returned, even if error is non-nil. If a parse error occurs, the message is
 // returned as application/octet-stream, and headers can still be read if they
 // were valid.
-func EnsurePart(r io.ReaderAt, size int64) (Part, error) {
-	p, err := Parse(r)
+//
+// If strict is set, fewer attempts are made to continue parsing when errors are
+// encountered, such as with invalid content-type headers or bare carriage returns.
+func EnsurePart(log *mlog.Log, strict bool, r io.ReaderAt, size int64) (Part, error) {
+	p, err := Parse(log, strict, r)
 	if err == nil {
-		err = p.Walk(nil)
+		err = p.Walk(log, nil)
 	}
 	if err != nil {
 		np, err2 := fallbackPart(p, r, size)
@@ -183,7 +187,7 @@ func (p *Part) SetMessageReaderAt() error {
 }
 
 // Walk through message, decoding along the way, and collecting mime part offsets and sizes, and line counts.
-func (p *Part) Walk(parent *Part) error {
+func (p *Part) Walk(log *mlog.Log, parent *Part) error {
 	if len(p.bound) == 0 {
 		if p.MediaType == "MESSAGE" && (p.MediaSubType == "RFC822" || p.MediaSubType == "GLOBAL") {
 			// todo: don't read whole submessage in memory...
@@ -192,11 +196,11 @@ func (p *Part) Walk(parent *Part) error {
 				return err
 			}
 			br := bytes.NewReader(buf)
-			mp, err := Parse(br)
+			mp, err := Parse(log, p.strict, br)
 			if err != nil {
 				return fmt.Errorf("parsing embedded message: %w", err)
 			}
-			if err := mp.Walk(nil); err != nil {
+			if err := mp.Walk(log, nil); err != nil {
 				// If this is a DSN and we are not in pedantic mode, accept unexpected end of
 				// message. This is quite common because MTA's sometimes just truncate the original
 				// message in a place that makes the message invalid.
@@ -218,14 +222,14 @@ func (p *Part) Walk(parent *Part) error {
 	}
 
 	for {
-		pp, err := p.ParseNextPart()
+		pp, err := p.ParseNextPart(log)
 		if err == io.EOF {
 			return nil
 		}
 		if err != nil {
 			return err
 		}
-		if err := pp.Walk(p); err != nil {
+		if err := pp.Walk(log, p); err != nil {
 			return err
 		}
 	}
@@ -239,7 +243,7 @@ func (p *Part) String() string {
 // newPart parses a new part, which can be the top-level message.
 // offset is the bound offset for parts, and the start of message for top-level messages. parent indicates if this is a top-level message or sub-part.
 // If an error occurs, p's exported values can still be relevant. EnsurePart uses these values.
-func newPart(r io.ReaderAt, offset int64, parent *Part) (p Part, rerr error) {
+func newPart(log *mlog.Log, strict bool, r io.ReaderAt, offset int64, parent *Part) (p Part, rerr error) {
 	if r == nil {
 		panic("nil reader")
 	}
@@ -248,9 +252,10 @@ func newPart(r io.ReaderAt, offset int64, parent *Part) (p Part, rerr error) {
 		EndOffset:      -1,
 		r:              r,
 		parent:         parent,
+		strict:         strict,
 	}
 
-	b := &bufAt{r: r, offset: offset}
+	b := &bufAt{strict: strict, r: r, offset: offset}
 
 	if parent != nil {
 		p.BoundaryOffset = offset
@@ -297,16 +302,46 @@ func newPart(r io.ReaderAt, offset int64, parent *Part) (p Part, rerr error) {
 	ct := p.header.Get("Content-Type")
 	mt, params, err := mime.ParseMediaType(ct)
 	if err != nil && ct != "" {
-		return p, fmt.Errorf("%w: %s: %q", ErrBadContentType, err, ct)
-	}
-	if mt != "" {
+		if moxvar.Pedantic || strict {
+			return p, fmt.Errorf("%w: %s: %q", ErrBadContentType, err, ct)
+		}
+
+		// Try parsing just a content-type, ignoring parameters.
+		// ../rfc/2045:628
+		ct = strings.TrimSpace(strings.SplitN(ct, ";", 2)[0])
+		t := strings.SplitN(ct, "/", 2)
+		isToken := func(s string) bool {
+			const separators = `()<>@,;:\\"/[]?= ` // ../rfc/2045:663
+			for _, c := range s {
+				if c < 0x20 || c >= 0x80 || strings.ContainsRune(separators, c) {
+					return false
+				}
+			}
+			return len(s) > 0
+		}
+		// We cannot recover content-type of multipart, we won't have a boundary.
+		if len(t) == 2 && isToken(t[0]) && !strings.EqualFold(t[0], "multipart") && isToken(t[1]) {
+			p.MediaType = strings.ToUpper(t[0])
+			p.MediaSubType = strings.ToUpper(t[1])
+		} else {
+			p.MediaType = "APPLICATION"
+			p.MediaSubType = "OCTET-STREAM"
+		}
+		log.Debugx("malformed content-type, attempting to recover and continuing", err, mlog.Field("contenttype", p.header.Get("Content-Type")), mlog.Field("mediatype", p.MediaType), mlog.Field("mediasubtype", p.MediaSubType))
+	} else if mt != "" {
 		t := strings.SplitN(strings.ToUpper(mt), "/", 2)
 		if len(t) != 2 {
-			return p, fmt.Errorf("bad content-type: %q (content-type %q)", mt, ct)
+			if moxvar.Pedantic || strict {
+				return p, fmt.Errorf("bad content-type: %q (content-type %q)", mt, ct)
+			}
+			log.Debug("malformed media-type, ignoring and continuing", mlog.Field("type", mt))
+			p.MediaType = "APPLICATION"
+			p.MediaSubType = "OCTET-STREAM"
+		} else {
+			p.MediaType = t[0]
+			p.MediaSubType = t[1]
+			p.ContentTypeParams = params
 		}
-		p.MediaType = t[0]
-		p.MediaSubType = t[1]
-		p.ContentTypeParams = params
 	}
 
 	p.ContentID = p.header.Get("Content-Id")
@@ -314,7 +349,7 @@ func newPart(r io.ReaderAt, offset int64, parent *Part) (p Part, rerr error) {
 	p.ContentTransferEncoding = strings.ToUpper(p.header.Get("Content-Transfer-Encoding"))
 
 	if parent == nil {
-		p.Envelope, err = parseEnvelope(mail.Header(p.header))
+		p.Envelope, err = parseEnvelope(log, mail.Header(p.header))
 		if err != nil {
 			return p, err
 		}
@@ -411,7 +446,7 @@ var wordDecoder = mime.WordDecoder{
 	},
 }
 
-func parseEnvelope(h mail.Header) (*Envelope, error) {
+func parseEnvelope(log *mlog.Log, h mail.Header) (*Envelope, error) {
 	date, _ := h.Date()
 
 	// We currently marshal this field to JSON. But JSON cannot represent all
@@ -433,19 +468,19 @@ func parseEnvelope(h mail.Header) (*Envelope, error) {
 	env := &Envelope{
 		date,
 		subject,
-		parseAddressList(h, "from"),
-		parseAddressList(h, "sender"),
-		parseAddressList(h, "reply-to"),
-		parseAddressList(h, "to"),
-		parseAddressList(h, "cc"),
-		parseAddressList(h, "bcc"),
+		parseAddressList(log, h, "from"),
+		parseAddressList(log, h, "sender"),
+		parseAddressList(log, h, "reply-to"),
+		parseAddressList(log, h, "to"),
+		parseAddressList(log, h, "cc"),
+		parseAddressList(log, h, "bcc"),
 		h.Get("In-Reply-To"),
 		h.Get("Message-Id"),
 	}
 	return env, nil
 }
 
-func parseAddressList(h mail.Header, k string) []Address {
+func parseAddressList(log *mlog.Log, h mail.Header, k string) []Address {
 	l, err := h.AddressList(k)
 	if err != nil {
 		return nil
@@ -457,7 +492,7 @@ func parseAddressList(h mail.Header, k string) []Address {
 		addr, err := smtp.ParseAddress(a.Address)
 		if err != nil {
 			// todo: pass a ctx to this function so we can log with cid.
-			xlog.Infox("parsing address", err, mlog.Field("address", a.Address))
+			log.Infox("parsing address (continuing)", err, mlog.Field("address", a.Address))
 		} else {
 			user = addr.Localpart.String()
 			host = addr.Domain.ASCII
@@ -470,7 +505,7 @@ func parseAddressList(h mail.Header, k string) []Address {
 // ParseNextPart parses the next (sub)part of this multipart message.
 // ParseNextPart returns io.EOF and a nil part when there are no more parts.
 // Only used for initial parsing of message. Once parsed, use p.Parts.
-func (p *Part) ParseNextPart() (*Part, error) {
+func (p *Part) ParseNextPart(log *mlog.Log) (*Part, error) {
 	if len(p.bound) == 0 {
 		return nil, errNotMultipart
 	}
@@ -479,7 +514,7 @@ func (p *Part) ParseNextPart() (*Part, error) {
 			panic("access not sequential")
 		}
 		// Set nextBoundOffset by fully reading the last part.
-		last, err := newPart(p.r, p.lastBoundOffset, p)
+		last, err := newPart(log, p.strict, p.r, p.lastBoundOffset, p)
 		if err != nil {
 			return nil, err
 		}
@@ -490,7 +525,7 @@ func (p *Part) ParseNextPart() (*Part, error) {
 			return nil, fmt.Errorf("internal error: reading part did not set nextBoundOffset")
 		}
 	}
-	b := &bufAt{r: p.r, offset: p.nextBoundOffset}
+	b := &bufAt{strict: p.strict, r: p.r, offset: p.nextBoundOffset}
 	// todo: should we require a crlf on final closing bound? we don't require it because some message/rfc822 don't have a crlf after their closing boundary, so those messages don't end in crlf.
 	line, crlf, err := b.ReadLine(false)
 	if err != nil {
@@ -523,7 +558,7 @@ func (p *Part) ParseNextPart() (*Part, error) {
 	boundOffset := p.nextBoundOffset
 	p.lastBoundOffset = boundOffset
 	p.nextBoundOffset = -1
-	np, err := newPart(p.r, boundOffset, p)
+	np, err := newPart(log, p.strict, p.r, boundOffset, p)
 	if err != nil {
 		return nil, err
 	}
@@ -623,13 +658,37 @@ func (p *Part) RawReader() io.Reader {
 		panic("missing reader")
 	}
 	if p.EndOffset >= 0 {
-		return io.NewSectionReader(p.r, p.BodyOffset, p.EndOffset-p.BodyOffset)
+		return &crlfReader{strict: p.strict, r: io.NewSectionReader(p.r, p.BodyOffset, p.EndOffset-p.BodyOffset)}
 	}
 	p.RawLineCount = 0
 	if p.parent == nil {
-		return &offsetReader{p, p.BodyOffset, true, false}
+		return &offsetReader{p, p.BodyOffset, p.strict, true, false}
 	}
-	return &boundReader{p: p, b: &bufAt{r: p.r, offset: p.BodyOffset}, prevlf: true}
+	return &boundReader{p: p, b: &bufAt{strict: p.strict, r: p.r, offset: p.BodyOffset}, prevlf: true}
+}
+
+// crlfReader verifies there are no bare newlines and optionally no bare carriage returns.
+type crlfReader struct {
+	r      io.Reader
+	strict bool
+	prevcr bool
+}
+
+func (r *crlfReader) Read(buf []byte) (int, error) {
+	n, err := r.r.Read(buf)
+	if err == nil || err == io.EOF {
+		for _, b := range buf[:n] {
+			if b == '\n' && !r.prevcr {
+				err = errBareLF
+				break
+			} else if b != '\n' && r.prevcr && (r.strict || moxvar.Pedantic) {
+				err = errBareCR
+				break
+			}
+			r.prevcr = b == '\r'
+		}
+	}
+	return n, err
 }
 
 // bufAt is a buffered reader on an underlying ReaderAt.
@@ -637,6 +696,7 @@ func (p *Part) RawReader() io.Reader {
 type bufAt struct {
 	offset int64 // Offset in r currently consumed, i.e. not including any buffered data.
 
+	strict  bool
 	r       io.ReaderAt
 	buf     []byte // Buffered data.
 	nbuf    int    // Valid bytes in buf.
@@ -698,7 +758,7 @@ func (b *bufAt) line(consume, requirecrlf bool) (buf []byte, crlf bool, err erro
 		}
 		i++
 		if i >= b.nbuf || b.buf[i] != '\n' {
-			if moxvar.Pedantic {
+			if b.strict || moxvar.Pedantic {
 				return nil, false, errBareCR
 			}
 			continue
@@ -743,6 +803,7 @@ func (b *bufAt) PeekByte() (byte, error) {
 type offsetReader struct {
 	p      *Part
 	offset int64
+	strict bool
 	prevlf bool
 	prevcr bool
 }
@@ -759,7 +820,7 @@ func (r *offsetReader) Read(buf []byte) (int, error) {
 			if err == nil || err == io.EOF {
 				if c == '\n' && !r.prevcr {
 					err = errBareLF
-				} else if c != '\n' && r.prevcr && moxvar.Pedantic {
+				} else if c != '\n' && r.prevcr && (r.strict || moxvar.Pedantic) {
 					err = errBareCR
 				}
 			}
@@ -784,7 +845,6 @@ type boundReader struct {
 	nbuf   int    // Number of valid bytes in buf.
 	crlf   []byte // Possible crlf, to be returned if we do not yet encounter a boundary.
 	prevlf bool   // If last char returned was a newline. For counting lines.
-	prevcr bool
 }
 
 func (b *boundReader) Read(buf []byte) (count int, rerr error) {
@@ -795,15 +855,7 @@ func (b *boundReader) Read(buf []byte) (count int, rerr error) {
 				if b.prevlf {
 					b.p.RawLineCount++
 				}
-				if rerr == nil || rerr == io.EOF {
-					if c == '\n' && !b.prevcr {
-						rerr = errBareLF
-					} else if c != '\n' && b.prevcr && moxvar.Pedantic {
-						rerr = errBareCR
-					}
-				}
 				b.prevlf = c == '\n'
-				b.prevcr = c == '\r'
 			}
 		}
 	}()
diff --git a/message/part_test.go b/message/part_test.go
index 9e92460..0d19ff9 100644
--- a/message/part_test.go
+++ b/message/part_test.go
@@ -11,9 +11,12 @@ import (
 	"strings"
 	"testing"
 
+	"github.com/mjl-/mox/mlog"
 	"github.com/mjl-/mox/moxvar"
 )
 
+var xlog = mlog.New("message")
+
 func tcheck(t *testing.T, err error, msg string) {
 	t.Helper()
 	if err != nil {
@@ -37,7 +40,7 @@ func tfail(t *testing.T, err, expErr error) {
 
 func TestEmptyHeader(t *testing.T) {
 	s := "\r\nx"
-	p, err := EnsurePart(strings.NewReader(s), int64(len(s)))
+	p, err := EnsurePart(xlog, true, strings.NewReader(s), int64(len(s)))
 	tcheck(t, err, "parse empty headers")
 	buf, err := io.ReadAll(p.Reader())
 	tcheck(t, err, "read")
@@ -48,15 +51,85 @@ func TestEmptyHeader(t *testing.T) {
 }
 
 func TestBadContentType(t *testing.T) {
+	expBody := "test"
+
+	// Pedantic is like strict.
+	moxvar.Pedantic = true
 	s := "content-type: text/html;;\r\n\r\ntest"
-	p, err := EnsurePart(strings.NewReader(s), int64(len(s)))
+	p, err := EnsurePart(xlog, false, strings.NewReader(s), int64(len(s)))
 	tfail(t, err, ErrBadContentType)
 	buf, err := io.ReadAll(p.Reader())
 	tcheck(t, err, "read")
-	expBody := "test"
 	tcompare(t, string(buf), expBody)
 	tcompare(t, p.MediaType, "APPLICATION")
 	tcompare(t, p.MediaSubType, "OCTET-STREAM")
+	moxvar.Pedantic = false
+
+	// Strict
+	s = "content-type: text/html;;\r\n\r\ntest"
+	p, err = EnsurePart(xlog, true, strings.NewReader(s), int64(len(s)))
+	tfail(t, err, ErrBadContentType)
+	buf, err = io.ReadAll(p.Reader())
+	tcheck(t, err, "read")
+	tcompare(t, string(buf), expBody)
+	tcompare(t, p.MediaType, "APPLICATION")
+	tcompare(t, p.MediaSubType, "OCTET-STREAM")
+
+	// Non-strict but unrecoverable content-type.
+	s = "content-type: not a content type;;\r\n\r\ntest"
+	p, err = EnsurePart(xlog, false, strings.NewReader(s), int64(len(s)))
+	tcheck(t, err, "parsing message with bad but recoverable content-type")
+	buf, err = io.ReadAll(p.Reader())
+	tcheck(t, err, "read")
+	tcompare(t, string(buf), expBody)
+	tcompare(t, p.MediaType, "APPLICATION")
+	tcompare(t, p.MediaSubType, "OCTET-STREAM")
+
+	// We try to use only the content-type, typically better than application/octet-stream.
+	s = "content-type: text/html;;\r\n\r\ntest"
+	p, err = EnsurePart(xlog, false, strings.NewReader(s), int64(len(s)))
+	tcheck(t, err, "parsing message with bad but recoverable content-type")
+	buf, err = io.ReadAll(p.Reader())
+	tcheck(t, err, "read")
+	tcompare(t, string(buf), expBody)
+	tcompare(t, p.MediaType, "TEXT")
+	tcompare(t, p.MediaSubType, "HTML")
+
+	// Not recovering multipart, we won't have a boundary.
+	s = "content-type: multipart/mixed;;\r\n\r\ntest"
+	p, err = EnsurePart(xlog, false, strings.NewReader(s), int64(len(s)))
+	tcheck(t, err, "parsing message with bad but recoverable content-type")
+	buf, err = io.ReadAll(p.Reader())
+	tcheck(t, err, "read")
+	tcompare(t, string(buf), expBody)
+	tcompare(t, p.MediaType, "APPLICATION")
+	tcompare(t, p.MediaSubType, "OCTET-STREAM")
+}
+
+func TestBareCR(t *testing.T) {
+	s := "content-type: text/html\r\n\r\nbare\rcr\r\n"
+	expBody := "bare\rcr\r\n"
+
+	// Pedantic is like strict.
+	moxvar.Pedantic = true
+	p, err := EnsurePart(xlog, false, strings.NewReader(s), int64(len(s)))
+	tfail(t, err, errBareCR)
+	_, err = io.ReadAll(p.Reader())
+	tfail(t, err, errBareCR)
+	moxvar.Pedantic = false
+
+	// Strict.
+	p, err = EnsurePart(xlog, true, strings.NewReader(s), int64(len(s)))
+	tfail(t, err, errBareCR)
+	_, err = io.ReadAll(p.Reader())
+	tcheck(t, err, "read fallback part without error")
+
+	// Non-strict allows bare cr.
+	p, err = EnsurePart(xlog, false, strings.NewReader(s), int64(len(s)))
+	tcheck(t, err, "parse")
+	buf, err := io.ReadAll(p.Reader())
+	tcheck(t, err, "read")
+	tcompare(t, string(buf), expBody)
 }
 
 var basicMsg = strings.ReplaceAll(`From: <mjl@mox.example>
@@ -68,7 +141,7 @@ aGkK
 
 func TestBasic(t *testing.T) {
 	r := strings.NewReader(basicMsg)
-	p, err := Parse(r)
+	p, err := Parse(xlog, true, r)
 	tcheck(t, err, "new reader")
 
 	buf, err := io.ReadAll(p.RawReader())
@@ -103,7 +176,7 @@ Hello Joe, do you think we can meet at 3:30 tomorrow?
 
 func TestBasic2(t *testing.T) {
 	r := strings.NewReader(basicMsg2)
-	p, err := Parse(r)
+	p, err := Parse(xlog, true, r)
 	tcheck(t, err, "new reader")
 
 	buf, err := io.ReadAll(p.RawReader())
@@ -123,9 +196,9 @@ func TestBasic2(t *testing.T) {
 	}
 
 	r = strings.NewReader(basicMsg2)
-	p, err = Parse(r)
+	p, err = Parse(xlog, true, r)
 	tcheck(t, err, "new reader")
-	err = p.Walk(nil)
+	err = p.Walk(xlog, nil)
 	tcheck(t, err, "walk")
 	if p.RawLineCount != 2 {
 		t.Fatalf("basic message, got %d lines, expected 2", p.RawLineCount)
@@ -164,25 +237,25 @@ This is the epilogue.  It is also to be ignored.
 func TestMime(t *testing.T) {
 	// from ../rfc/2046:1148
 	r := strings.NewReader(mimeMsg)
-	p, err := Parse(r)
+	p, err := Parse(xlog, true, r)
 	tcheck(t, err, "new reader")
 	if len(p.bound) == 0 {
 		t.Fatalf("got no bound, expected bound for mime message")
 	}
 
-	pp, err := p.ParseNextPart()
+	pp, err := p.ParseNextPart(xlog)
 	tcheck(t, err, "next part")
 	buf, err := io.ReadAll(pp.Reader())
 	tcheck(t, err, "read all")
 	tcompare(t, string(buf), "This is implicitly typed plain US-ASCII text.\r\nIt does NOT end with a linebreak.")
 
-	pp, err = p.ParseNextPart()
+	pp, err = p.ParseNextPart(xlog)
 	tcheck(t, err, "next part")
 	buf, err = io.ReadAll(pp.Reader())
 	tcheck(t, err, "read all")
 	tcompare(t, string(buf), "This is explicitly typed plain US-ASCII text.\r\nIt DOES end with a linebreak.\r\n")
 
-	_, err = p.ParseNextPart()
+	_, err = p.ParseNextPart(xlog)
 	tcompare(t, err, io.EOF)
 
 	if len(p.Parts) != 2 {
@@ -201,33 +274,38 @@ func TestLongLine(t *testing.T) {
 	for i := range line {
 		line[i] = 'a'
 	}
-	_, err := Parse(bytes.NewReader(line))
+	_, err := Parse(xlog, true, bytes.NewReader(line))
 	tfail(t, err, errLineTooLong)
 }
 
-func TestHalfCrLf(t *testing.T) {
-	parse := func(s string) error {
-		p, err := Parse(strings.NewReader(s))
+func TestBareCrLf(t *testing.T) {
+	parse := func(strict bool, s string) error {
+		p, err := Parse(xlog, strict, strings.NewReader(s))
 		if err != nil {
 			return err
 		}
-		return p.Walk(nil)
+		return p.Walk(xlog, nil)
 	}
-	err := parse("subject: test\ntest\r\n")
+	err := parse(false, "subject: test\ntest\r\n")
 	tfail(t, err, errBareLF)
-	err = parse("\r\ntest\ntest\r\n")
+	err = parse(false, "\r\ntest\ntest\r\n")
 	tfail(t, err, errBareLF)
 
 	moxvar.Pedantic = true
-	err = parse("subject: test\rtest\r\n")
+	err = parse(false, "subject: test\rtest\r\n")
 	tfail(t, err, errBareCR)
-	err = parse("\r\ntest\rtest\r\n")
+	err = parse(false, "\r\ntest\rtest\r\n")
 	tfail(t, err, errBareCR)
 	moxvar.Pedantic = false
 
-	err = parse("subject: test\rtest\r\n")
+	err = parse(true, "subject: test\rtest\r\n")
+	tfail(t, err, errBareCR)
+	err = parse(true, "\r\ntest\rtest\r\n")
+	tfail(t, err, errBareCR)
+
+	err = parse(false, "subject: test\rtest\r\n")
 	tcheck(t, err, "header with bare cr")
-	err = parse("\r\ntest\rtest\r\n")
+	err = parse(false, "\r\ntest\rtest\r\n")
 	tcheck(t, err, "body with bare cr")
 }
 
@@ -238,25 +316,25 @@ func TestMissingClosingBoundary(t *testing.T) {
 
 test
 `, "\n", "\r\n")
-	msg, err := Parse(strings.NewReader(message))
+	msg, err := Parse(xlog, false, strings.NewReader(message))
 	tcheck(t, err, "new reader")
 	err = walkmsg(&msg)
 	tfail(t, err, errMissingClosingBoundary)
 
-	msg, _ = Parse(strings.NewReader(message))
-	err = msg.Walk(nil)
+	msg, _ = Parse(xlog, false, strings.NewReader(message))
+	err = msg.Walk(xlog, nil)
 	tfail(t, err, errMissingClosingBoundary)
 }
 
 func TestHeaderEOF(t *testing.T) {
 	message := "header: test"
-	_, err := Parse(strings.NewReader(message))
+	_, err := Parse(xlog, false, strings.NewReader(message))
 	tfail(t, err, errUnexpectedEOF)
 }
 
 func TestBodyEOF(t *testing.T) {
 	message := "header: test\r\n\r\ntest"
-	msg, err := Parse(strings.NewReader(message))
+	msg, err := Parse(xlog, true, strings.NewReader(message))
 	tcheck(t, err, "new reader")
 	buf, err := io.ReadAll(msg.Reader())
 	tcheck(t, err, "read body")
@@ -287,7 +365,7 @@ test
 
 `, "\n", "\r\n")
 
-	msg, err := Parse(strings.NewReader(message))
+	msg, err := Parse(xlog, false, strings.NewReader(message))
 	tcheck(t, err, "new reader")
 	enforceSequential = true
 	defer func() {
@@ -296,8 +374,8 @@ test
 	err = walkmsg(&msg)
 	tcheck(t, err, "walkmsg")
 
-	msg, _ = Parse(strings.NewReader(message))
-	err = msg.Walk(nil)
+	msg, _ = Parse(xlog, false, strings.NewReader(message))
+	err = msg.Walk(xlog, nil)
 	tcheck(t, err, "msg.Walk")
 }
 
@@ -374,7 +452,7 @@ Content-Transfer-Encoding: Quoted-printable
 --unique-boundary-1--
 `, "\n", "\r\n")
 
-	msg, err := Parse(strings.NewReader(nestedMessage))
+	msg, err := Parse(xlog, true, strings.NewReader(nestedMessage))
 	tcheck(t, err, "new reader")
 	enforceSequential = true
 	defer func() {
@@ -399,8 +477,8 @@ Content-Transfer-Encoding: Quoted-printable
 		t.Fatalf("got %q, expected %q", buf, exp)
 	}
 
-	msg, _ = Parse(strings.NewReader(nestedMessage))
-	err = msg.Walk(nil)
+	msg, _ = Parse(xlog, false, strings.NewReader(nestedMessage))
+	err = msg.Walk(xlog, nil)
 	tcheck(t, err, "msg.Walk")
 
 }
@@ -440,7 +518,7 @@ func walk(path string) error {
 		return err
 	}
 	defer r.Close()
-	msg, err := Parse(r)
+	msg, err := Parse(xlog, false, r)
 	if err != nil {
 		return err
 	}
@@ -460,7 +538,7 @@ func walkmsg(msg *Part) error {
 		}
 
 		if msg.MediaType == "MESSAGE" && (msg.MediaSubType == "RFC822" || msg.MediaSubType == "GLOBAL") {
-			mp, err := Parse(bytes.NewReader(buf))
+			mp, err := Parse(xlog, false, bytes.NewReader(buf))
 			if err != nil {
 				return err
 			}
@@ -488,7 +566,7 @@ func walkmsg(msg *Part) error {
 	}
 
 	for {
-		pp, err := msg.ParseNextPart()
+		pp, err := msg.ParseNextPart(xlog)
 		if err == io.EOF {
 			return nil
 		}
@@ -507,7 +585,7 @@ func TestEmbedded(t *testing.T) {
 	tcheck(t, err, "open")
 	fi, err := f.Stat()
 	tcheck(t, err, "stat")
-	_, err = EnsurePart(f, fi.Size())
+	_, err = EnsurePart(xlog, false, f, fi.Size())
 	tcheck(t, err, "parse")
 }
 
@@ -516,6 +594,6 @@ func TestEmbedded2(t *testing.T) {
 	tcheck(t, err, "readfile")
 	buf = bytes.ReplaceAll(buf, []byte("\n"), []byte("\r\n"))
 
-	_, err = EnsurePart(bytes.NewReader(buf), int64(len(buf)))
+	_, err = EnsurePart(xlog, false, bytes.NewReader(buf), int64(len(buf)))
 	tfail(t, err, nil)
 }
diff --git a/smtpserver/analyze.go b/smtpserver/analyze.go
index 3c05867..294e010 100644
--- a/smtpserver/analyze.go
+++ b/smtpserver/analyze.go
@@ -141,7 +141,7 @@ func analyze(ctx context.Context, log *mlog.Log, resolver dns.Resolver, d delive
 		// Messages with DMARC aggregate reports must have a dmarc pass. ../rfc/7489:1866
 		if d.dmarcResult.Status != dmarc.StatusPass {
 			log.Info("received dmarc report without dmarc pass, not processing as dmarc report")
-		} else if report, err := dmarcrpt.ParseMessageReport(store.FileMsgReader(d.m.MsgPrefix, d.dataFile)); err != nil {
+		} else if report, err := dmarcrpt.ParseMessageReport(log, store.FileMsgReader(d.m.MsgPrefix, d.dataFile)); err != nil {
 			log.Infox("parsing dmarc report", err)
 		} else if d, err := dns.ParseDomain(report.PolicyPublished.Domain); err != nil {
 			log.Infox("parsing domain in dmarc report", err)
@@ -173,7 +173,7 @@ func analyze(ctx context.Context, log *mlog.Log, resolver dns.Resolver, d delive
 
 		if !ok {
 			log.Info("received mail to tlsrpt without acceptable DKIM signature, not processing as tls report")
-		} else if report, err := tlsrpt.ParseMessage(store.FileMsgReader(d.m.MsgPrefix, d.dataFile)); err != nil {
+		} else if report, err := tlsrpt.ParseMessage(log, store.FileMsgReader(d.m.MsgPrefix, d.dataFile)); err != nil {
 			log.Infox("parsing tls report", err)
 		} else {
 			var known bool
@@ -274,7 +274,7 @@ func analyze(ctx context.Context, log *mlog.Log, resolver dns.Resolver, d delive
 			log.Errorx("get key for verifying subject token", err)
 			return reject(smtp.C451LocalErr, smtp.SeSys3Other0, "error processing", err, reasonSubjectpassError)
 		}
-		err = subjectpass.Verify(d.dataFile, []byte(subjectpassKey), conf.SubjectPass.Period)
+		err = subjectpass.Verify(log, d.dataFile, []byte(subjectpassKey), conf.SubjectPass.Period)
 		pass := err == nil
 		log.Infox("pass by subject token", err, mlog.Field("pass", pass))
 		if pass {
diff --git a/smtpserver/rejects.go b/smtpserver/rejects.go
index 45ef6b7..0ff9273 100644
--- a/smtpserver/rejects.go
+++ b/smtpserver/rejects.go
@@ -18,7 +18,7 @@ import (
 
 // rejectPresent returns whether the message is already present in the rejects mailbox.
 func rejectPresent(log *mlog.Log, acc *store.Account, rejectsMailbox string, m *store.Message, f *os.File) (present bool, msgID string, hash []byte, rerr error) {
-	if p, err := message.Parse(store.FileMsgReader(m.MsgPrefix, f)); err != nil {
+	if p, err := message.Parse(log, false, store.FileMsgReader(m.MsgPrefix, f)); err != nil {
 		log.Infox("parsing reject message for message-id", err)
 	} else if header, err := p.Header(); err != nil {
 		log.Infox("parsing reject message header for message-id", err)
diff --git a/smtpserver/server.go b/smtpserver/server.go
index f23ba9b..b12c063 100644
--- a/smtpserver/server.go
+++ b/smtpserver/server.go
@@ -1548,9 +1548,9 @@ func (c *conn) cmdData(p *parser) {
 
 	if Localserve {
 		// Require that message can be parsed fully.
-		p, err := message.Parse(dataFile)
+		p, err := message.Parse(c.log, false, dataFile)
 		if err == nil {
-			err = p.Walk(nil)
+			err = p.Walk(c.log, nil)
 		}
 		if err != nil {
 			// ../rfc/6409:541
@@ -1661,7 +1661,7 @@ func (c *conn) submit(ctx context.Context, recvHdrFor func(string) string, msgWr
 	// for other users.
 	// We don't check the Sender field, there is no expectation of verification, ../rfc/7489:2948
 	// and with Resent headers it seems valid to have someone else as Sender. ../rfc/5322:1578
-	msgFrom, header, err := message.From(dataFile)
+	msgFrom, header, err := message.From(c.log, true, dataFile)
 	if err != nil {
 		metricSubmission.WithLabelValues("badmessage").Inc()
 		c.log.Infox("parsing message From address", err, mlog.Field("user", c.username))
@@ -1854,7 +1854,7 @@ func (c *conn) deliver(ctx context.Context, recvHdrFor func(string) string, msgW
 
 	// todo: in decision making process, if we run into (some) temporary errors, attempt to continue. if we decide to accept, all good. if we decide to reject, we'll make it a temporary reject.
 
-	msgFrom, headers, err := message.From(dataFile)
+	msgFrom, headers, err := message.From(c.log, false, dataFile)
 	if err != nil {
 		c.log.Infox("parsing message for From address", err)
 	}
@@ -2388,7 +2388,7 @@ func (c *conn) deliver(ctx context.Context, recvHdrFor func(string) string, msgW
 
 		// Gather the message-id before we deliver and the file may be consumed.
 		if !parsedMessageID {
-			if p, err := message.Parse(store.FileMsgReader(m.MsgPrefix, dataFile)); err != nil {
+			if p, err := message.Parse(c.log, false, store.FileMsgReader(m.MsgPrefix, dataFile)); err != nil {
 				log.Infox("parsing message for message-id", err)
 			} else if header, err := p.Header(); err != nil {
 				log.Infox("parsing message header for message-id", err)
diff --git a/store/account.go b/store/account.go
index aecea0d..559c014 100644
--- a/store/account.go
+++ b/store/account.go
@@ -1045,7 +1045,7 @@ func (a *Account) DeliverMessage(log *mlog.Log, tx *bstore.Tx, m *Message, msgFi
 	var part *message.Part
 	if m.ParsedBuf == nil {
 		mr := FileMsgReader(m.MsgPrefix, msgFile) // We don't close, it would close the msgFile.
-		p, err := message.EnsurePart(mr, m.Size)
+		p, err := message.EnsurePart(log, false, mr, m.Size)
 		if err != nil {
 			log.Infox("parsing delivered message", err, mlog.Field("parse", ""), mlog.Field("message", m.ID))
 			// We continue, p is still valid.
@@ -1365,7 +1365,7 @@ func MessageRuleset(log *mlog.Log, dest config.Destination, m *Message, msgPrefi
 	}
 
 	mr := FileMsgReader(msgPrefix, msgFile) // We don't close, it would close the msgFile.
-	p, err := message.Parse(mr)
+	p, err := message.Parse(log, false, mr)
 	if err != nil {
 		log.Errorx("parsing message for evaluating rulesets, continuing with headers", err, mlog.Field("parse", ""))
 		// note: part is still set.
diff --git a/subjectpass/subjectpass.go b/subjectpass/subjectpass.go
index dcb5958..5caa29b 100644
--- a/subjectpass/subjectpass.go
+++ b/subjectpass/subjectpass.go
@@ -76,7 +76,7 @@ func Generate(mailFrom smtp.Address, key []byte, tm time.Time) string {
 
 // Verify parses "message" and checks if it includes a subjectpass token in its
 // Subject header that is still valid (within "period") and signed with "key".
-func Verify(r io.ReaderAt, key []byte, period time.Duration) (rerr error) {
+func Verify(log *mlog.Log, r io.ReaderAt, key []byte, period time.Duration) (rerr error) {
 	var token string
 
 	defer func() {
@@ -89,7 +89,7 @@ func Verify(r io.ReaderAt, key []byte, period time.Duration) (rerr error) {
 		log.Debugx("subjectpass verify result", rerr, mlog.Field("token", token), mlog.Field("period", period))
 	}()
 
-	p, err := message.Parse(r)
+	p, err := message.Parse(log, true, r)
 	if err != nil {
 		return fmt.Errorf("%w: parse message: %s", ErrMessage, err)
 	}
diff --git a/subjectpass/subjectpass_test.go b/subjectpass/subjectpass_test.go
index b9948f7..a50a628 100644
--- a/subjectpass/subjectpass_test.go
+++ b/subjectpass/subjectpass_test.go
@@ -7,26 +7,29 @@ import (
 	"testing"
 	"time"
 
+	"github.com/mjl-/mox/mlog"
 	"github.com/mjl-/mox/smtp"
 )
 
+var xlog = mlog.New("subjectpass")
+
 func TestSubjectPass(t *testing.T) {
 	key := []byte("secret token")
 	addr, _ := smtp.ParseAddress("mox@mox.example")
 	sig := Generate(addr, key, time.Now())
 
 	message := fmt.Sprintf("From: <mox@mox.example>\r\nSubject: let me in %s\r\n\r\nthe message", sig)
-	if err := Verify(strings.NewReader(message), key, time.Hour); err != nil {
+	if err := Verify(xlog, strings.NewReader(message), key, time.Hour); err != nil {
 		t.Fatalf("verifyPassToken: %s", err)
 	}
 
-	if err := Verify(strings.NewReader(message), []byte("bad key"), time.Hour); err == nil {
+	if err := Verify(xlog, strings.NewReader(message), []byte("bad key"), time.Hour); err == nil {
 		t.Fatalf("verifyPassToken did not fail")
 	}
 
 	sig = Generate(addr, key, time.Now().Add(-time.Hour-257))
 	message = fmt.Sprintf("From: <mox@mox.example>\r\nSubject: let me in %s\r\n\r\nthe message", sig)
-	if err := Verify(strings.NewReader(message), key, time.Hour); !errors.Is(err, ErrExpired) {
+	if err := Verify(xlog, strings.NewReader(message), key, time.Hour); !errors.Is(err, ErrExpired) {
 		t.Fatalf("verifyPassToken should have expired")
 	}
 }
diff --git a/tlsrpt/report.go b/tlsrpt/report.go
index beb6e4a..d3bb891 100644
--- a/tlsrpt/report.go
+++ b/tlsrpt/report.go
@@ -10,6 +10,7 @@ import (
 	"time"
 
 	"github.com/mjl-/mox/message"
+	"github.com/mjl-/mox/mlog"
 	"github.com/mjl-/mox/moxio"
 )
 
@@ -137,9 +138,9 @@ func Parse(r io.Reader) (*Report, error) {
 // ParseMessage parses a Report from a mail message.
 // The maximum size of the message is 15MB, the maximum size of the
 // decompressed report is 20MB.
-func ParseMessage(r io.ReaderAt) (*Report, error) {
+func ParseMessage(log *mlog.Log, r io.ReaderAt) (*Report, error) {
 	// ../rfc/8460:905
-	p, err := message.Parse(&moxio.LimitAtReader{R: r, Limit: 15 * 1024 * 1024})
+	p, err := message.Parse(log, true, &moxio.LimitAtReader{R: r, Limit: 15 * 1024 * 1024})
 	if err != nil {
 		return nil, fmt.Errorf("parsing mail message: %s", err)
 	}
@@ -147,10 +148,10 @@ func ParseMessage(r io.ReaderAt) (*Report, error) {
 	// Using multipart appears optional, and similar to DMARC someone may decide to
 	// send it like that, so accept a report if it's the entire message.
 	const allow = true
-	return parseMessageReport(p, allow)
+	return parseMessageReport(log, p, allow)
 }
 
-func parseMessageReport(p message.Part, allow bool) (*Report, error) {
+func parseMessageReport(log *mlog.Log, p message.Part, allow bool) (*Report, error) {
 	if p.MediaType != "MULTIPART" {
 		if !allow {
 			return nil, ErrNoReport
@@ -159,7 +160,7 @@ func parseMessageReport(p message.Part, allow bool) (*Report, error) {
 	}
 
 	for {
-		sp, err := p.ParseNextPart()
+		sp, err := p.ParseNextPart(log)
 		if err == io.EOF {
 			return nil, ErrNoReport
 		}
@@ -169,7 +170,7 @@ func parseMessageReport(p message.Part, allow bool) (*Report, error) {
 		if p.MediaSubType == "REPORT" && p.ContentTypeParams["report-type"] != "tlsrpt" {
 			return nil, fmt.Errorf("unknown report-type parameter %q", p.ContentTypeParams["report-type"])
 		}
-		report, err := parseMessageReport(*sp, p.MediaSubType == "REPORT")
+		report, err := parseMessageReport(log, *sp, p.MediaSubType == "REPORT")
 		if err == ErrNoReport {
 			continue
 		} else if err != nil || report != nil {
diff --git a/tlsrpt/report_test.go b/tlsrpt/report_test.go
index e0102a2..f5f507f 100644
--- a/tlsrpt/report_test.go
+++ b/tlsrpt/report_test.go
@@ -109,19 +109,19 @@ func TestReport(t *testing.T) {
 		t.Fatalf("parsing report: %s", err)
 	}
 
-	if _, err := ParseMessage(strings.NewReader(tlsrptMessage)); err != nil {
+	if _, err := ParseMessage(xlog, strings.NewReader(tlsrptMessage)); err != nil {
 		t.Fatalf("parsing TLSRPT from message: %s", err)
 	}
 
-	if _, err := ParseMessage(strings.NewReader(tlsrptMessage2)); err != nil {
+	if _, err := ParseMessage(xlog, strings.NewReader(tlsrptMessage2)); err != nil {
 		t.Fatalf("parsing TLSRPT from message: %s", err)
 	}
 
-	if _, err := ParseMessage(strings.NewReader(strings.ReplaceAll(tlsrptMessage, "multipart/report", "multipart/related"))); err != ErrNoReport {
+	if _, err := ParseMessage(xlog, strings.NewReader(strings.ReplaceAll(tlsrptMessage, "multipart/report", "multipart/related"))); err != ErrNoReport {
 		t.Fatalf("got err %v, expected ErrNoReport", err)
 	}
 
-	if _, err := ParseMessage(strings.NewReader(strings.ReplaceAll(tlsrptMessage, "application/tlsrpt+json", "application/json"))); err != ErrNoReport {
+	if _, err := ParseMessage(xlog, strings.NewReader(strings.ReplaceAll(tlsrptMessage, "application/tlsrpt+json", "application/json"))); err != ErrNoReport {
 		t.Fatalf("got err %v, expected ErrNoReport", err)
 	}
 
@@ -134,7 +134,7 @@ func TestReport(t *testing.T) {
 		if err != nil {
 			t.Fatalf("open %q: %s", file, err)
 		}
-		if _, err := ParseMessage(f); err != nil {
+		if _, err := ParseMessage(xlog, f); err != nil {
 			t.Fatalf("parsing TLSRPT from message %q: %s", file.Name(), err)
 		}
 		f.Close()
@@ -144,6 +144,6 @@ func TestReport(t *testing.T) {
 func FuzzParseMessage(f *testing.F) {
 	f.Add(tlsrptMessage)
 	f.Fuzz(func(t *testing.T, s string) {
-		ParseMessage(strings.NewReader(s))
+		ParseMessage(xlog, strings.NewReader(s))
 	})
 }
diff --git a/tlsrptdb/db_test.go b/tlsrptdb/db_test.go
index 2e66a22..4e3d2a3 100644
--- a/tlsrptdb/db_test.go
+++ b/tlsrptdb/db_test.go
@@ -88,7 +88,7 @@ func TestReport(t *testing.T) {
 		if err != nil {
 			t.Fatalf("open %q: %s", file, err)
 		}
-		report, err := tlsrpt.ParseMessage(f)
+		report, err := tlsrpt.ParseMessage(xlog, f)
 		f.Close()
 		if err != nil {
 			t.Fatalf("parsing TLSRPT from message %q: %s", file.Name(), err)
diff --git a/webaccount/import.go b/webaccount/import.go
index 629778c..a886863 100644
--- a/webaccount/import.go
+++ b/webaccount/import.go
@@ -505,7 +505,7 @@ func importMessages(ctx context.Context, log *mlog.Log, token string, acc *store
 		}
 
 		// Parse message and store parsed information for later fast retrieval.
-		p, err := message.EnsurePart(f, m.Size)
+		p, err := message.EnsurePart(log, false, f, m.Size)
 		if err != nil {
 			problemf("parsing message %s: %s (continuing)", pos, err)
 		}