/ 
Imager-1.029
⭐ Starred 27 / Imager::LargeSamples
Security Advisories (3)
CVE-2026-8669 (2026-05-15)

Imager versions through 1.030 for Perl allow a heap out of bounds (OOB) write on crafted multi-frame GIF files. Imager::File::GIF's i_readgif_multi_low allocates a single per-row buffer GifRow sized for the GIF's global screen width 'SWidth' and reuses it across every image in the file. The page-match branch validates Image.Width + Image.Left > SWidth before each DGifGetLine write, but the parallel skip-image branch at imgif.c:790-805 calls DGifGetLine(GifFile, GifRow, Width) with no such check.

CVE-2026-13705 (2026-07-06)

Imager versions before 1.032 for Perl have a heap out-of-bounds read in the bundled Imager::File::SGI reader via a 16-bit RLE literal run in read_rgb_16_rle. read_rgb_16_rle guards each literal run with if (count > data_left), but count is a pixel count while every 16-bit sample consumes two bytes. The copy loop reads inp[0] * 256 + inp[1] and advances two bytes per pixel, so a run with data_left / 2 < count <= data_left passes the guard yet consumes 2 * count bytes and reads past the end of the buffer. The 8-bit path is unaffected because there one pixel is one byte. Reading a crafted SGI image through Imager->read triggers the over-read before the parser rejects the malformed image, which can crash the process.

CVE-2026-14454 (2026-07-08)

Imager versions before 1.033 for Perl treat unsigned EXIF IFD entry counts as signed. Imager mishandled large EXIF IFD entry count values, treating them as negative numbers. This could lead to an attempt to allocate a block nearly the size of the address space, which fails and kills the process. An attacker could craft an image with EXIF data that terminates a worker process.

NAME

Imager::LargeSamples - track/document large sample support

SYNOPSIS

# make a large sample image
my $im = Imager->new(..., bits => 16);

# call some method
my $result = $im->$method(...);

# was the image modified at its full sample size

DESCRIPTION

Imager has had in-memory support for large samples for years now, but many methods still don't work at the higher sample size when supplied with a large sample image.

This document will track which methods support large samples and which don't, for future improvements.

Support by method

Method       Support    Notes
------       -------    -----
arc          Partial    [1]
box          Partial    [2]
circle       Partial    [1]
convert      Full
copy         Full
crop         Full
difference   Full
filter       Partial    Depends on the filter.
flip         Full
flood_fill   Partial    [1]
getpixel     Full
getsamples   Full
getscanline  Full
map          None
masked       Full
matrix_transform
             Full
paste        Full
polygon      Partial    [1]
polyline     None
read         Partial    See L<File format large sample support>
read_multi   Partial    See L<File format large sample support>
rotate       Full
rubthrough   Full
scale        Partial    Some qtypes support large samples
scaleX       None
scaleY       None
setpixel     Full
setscanline  Full
string       Full       Preserves large samples, but most font drivers
                        generate 8 or fewer bits of levels of coverage.
transform    None
transform2   None
write        Partial    See L<File format large sample support>
write_multi  Partial    See L<File format large sample support>

[1] filling an area using the fill parameter works at the full depth of the image, using filled => 1 and color works at 8-bits/sample

[2] box() will fill the area at the supplied color, but outline at 8-bits/sample.

File format large sample support

  Format    Format samples    Imager support
  ------    --------------    --------------
  BMP       8                 8
  GIF       8                 8
  ICO       8                 8
  JPEG      8, 12             8
  PBM       1                 1
  PGM/PPM   1-16              read any, writes 8, 16
  PNG       1, 2, 4, 8, 16    1, 2, 4, 8 paletted
  	       	     	      1, 8, 16 gray (1 for is_monochrome() images)
			      8, 16 RGB
  RAW                         8
  SGI       8, 16             8, 16
  TGA       8                 8
  TIFF      (many)            read/write 8, 16, 32 contig rgb/grey images
                              read/write bi-level
                              read/write 4/8 paletted images

Filter larger sample support

Filter      Large sample support
------      --------------------
autolevels      No
bumpmap         No
bumpmap_complex No
contrast        No
conv            Yes
fountain        Yes
gaussian        Yes
gradgen         No
hardinvert      Yes
mosaic          No
postlevels      No
radnoise        No
turbnoise       No
unsharpmask     Yes
watermark       No

AUTHOR

Tony Cook <tonyc@cpan.org>